An approach to “Design for Reliability” in solid state lighting systems at high temperatures

doi:10.1016/j.microrel.2011.06.029

Microelectronics Reliability

Volume 52, Issue 5, May 2012, Pages 783-793

https://doi.org/10.1016/j.microrel.2011.06.029 Get rights and content

Abstract

Providing correct reliability information is critical for all manufacturers for both their customers and also their in-house departments intended for product improvements. Solid state lighting (SSL) technology as the novel lighting technology is not an exception. Due to the fact that SSL is a relatively new technology, research on its reliability issues and also systematic methods for design improvement is lacking.

In this paper, we introduce an approach for “Design for Reliability” in SSL. Our approach includes three major steps. The first step is the design phase where the physics of failure of the device and virtual assessment of the reliability of the device is investigated. Based on these results, their failure’s causes; and their relationship with reliability and lifetime of the SSL device is defined. Sensors are then assigned to the critical failure causes at the proper positions of the system. The second step of our approach is processing the data from sensors while device is functional. In this phase, gathered data from the sensors, are processed in order to calculate the lifetime of the device. The system should have a processing capability for accomplishing this task and also report the status to the maintenance system. The third step of our reliability approach is to have built-in self-maintenance capabilities. Therefore, the system can predict its failure and be functional for longer time.

At the second part of this paper, we introduced a SSL device which has been designed in our group which can fulfill the criteria for our “Design for Reliability approach”. This SSL device can provide around 200 lumen and has the capability of sensing and monitoring the high temperature in system. High junction temperature is one of the major failure causes in SSL devices. Due to the additional flexibility that smart controller of this device gives us, self-maintenance solution can be implemented in this device.

Highlights

► We introduce a methodology for SSL systems “Design for Reliability”. ► The methodology is based on stress and damage model of the device and monitoring the critical points. ► We design a smart SSL system and apply this methodology in it. ► The focus is on LED failure in high temperature and its junction temperature is the critical point. ► The system lifetime is predicted by monitoring LED junction temperature.

Introduction

This work focuses on reliability study and introduces an approach for “Design for Reliability” for SSL systems. During the initial stage of any reliability study, it is essential to know why and how this technology is used as reliability, or improvement in reliability is dependent on these parameters. Therefore, a very short introduction about the solid state lighting (SSL) technology will be presented in the following section.

The system architecture will be discussed in Section 2. As each parts of a system play an important role in terms of functionality and reliability of the system, a good knowledge on different components of the system is required. The main subject of Section 3 is about the reliability definition, the necessity of this study for SSL and finally how much research is already done in this field. In Section 4 our approach for “Design for Reliability” for SSL systems will be introduced. A flowchart of SSL starting from design phase till the end of the device useful life is the core of this section. In this flowchart, new steps in additional to the standard reliability test for SSL products are introduced for our “Design for Reliability” procedure. Our approach toward having “Design for Reliability” procedure in SSL life cycle will then be discussed. A complete smart SSL device has been designed within the framework of this study which can fulfill the criteria of our reliability approach. In Section 5, this system will be used as a example for our reliability approach.

Solid state lighting, commonly called SSL, is the new lighting technology based on light emitting diodes. Although LEDs has been used for a very long time for different applications (mostly as indicators), the idea of using LED for lighting applications has only been introduced in recent decade. The architecture of the first few lighting systems based on LEDs were using a large number (over 10 or 20) of 20 mA 3–5 V conventional LEDs to compensate for the required light output [1].

One of the main reason that that made the idea of using LED as the light source very attractive was its low power consumption with respect to the conventional lighting systems (Fig. 1). The electrical energy consumption for lighting in big cities is about 20–25% of the total consumed electrical energy. Therefore energy saving lighting technology is very appealing in our societies. Thus, lots of researches had been focusing on producing high power LEDs with high efficacy and better light quality, especially in the area of white LEDs for general purpose applications. Latest high power LEDs, which also known as high brightness LEDs (HB LEDs), has a power rating of 1 W and even higher. Recent SSL devices have remarkably high efficacy of up to 150 lm/W, as comparison to 15 lm/W for conventional 60–100 W incandescent light bulb; and 60 lm/W for compact florescent light [3].

Another advantage of SSL over conventional lighting systems is its long lifetime. A common incandescent lamp has an average lifetime of around 1000 h whereas the fluorescent lighting technology has an average lifetime of about 10 times longer than incandescent light which is around 10,000 h [1], [2]. However, fluorescent lighting technology is a mature technology now and the gain in longer lifetime has reached its limit and it will be very difficult for fluorescent lamps to have dramatically longer lifetime than 10,000 h in the near future. The present LED lighting system technologies have an average of lifetime around 20,000 h (this value varies from manufacturer to manufacturer) with the potential for its lifetime to reach up to 50,000 h and even more up to 100,000 h [4].

In conclusion, even though a single light bulb price of SSL still is much more expensive compared to a fluorescent light bulb price, the total lifetime cost of SSL systems will be much lower than the current lighting technologies considering the saving from the energy consumptions and longer lifetime of each bulb.

The third advantage of SSL lighting is that it is an environmental friendly technology. Fluorescent lighting systems which are one of the most efficient lighting systems before SSL systems, contains mercury which is not disposable. In contrary with fluorescent lighting devices, LEDs are semiconductor devices which are free of toxic materials. The dream of having a completely environmental friendly lighting system can be achieved by replacing the electronic components with lead-free devices.

SSL technology can be interesting for designers and consumers in lots of other ways. One example is the design flexibility that enables SSL technology to be quickly adopted into different lighting systems market. LEDs are semiconductor devices that their light intensity can be easily controlled. They can be in theory dimmed to 100% but their drivers’ technologies are still not completely compatible with current dimmers in the market. The SSL devices turn-on time is almost zero in contrary to the Fluorescent lighting system turn-on time which is more in the order of seconds. Furthermore, the ease for enabling the color changing capability of SSL system makes SSL appealing to special applications such as disco lighting or shopping center lighting. Due to narrow emission bandwidth of LEDs, there is no infra-red or ultra-violet light emission from SSL devices which is very important for a very good and sharp spot-lights [1], [3].

As mentioned in above discussion, there can be a wide variety of applications that SSL devices are very good replacements for traditional lighting systems. In-door lighting, spot lighting, street lighting and lighting of automotive applications are just some of the many examples. There are also lots of new applications which have been introduced to the market like decorative outdoor/indoor lighting. The examples of these new age lighting systems can be seen in lots of historical building, bridges and conference centers all around the world [5], [6].

Section snippets

Solid state lighting system architecture

Systems in SSL devices include three major parts; optical part, LED electrical driver and interconnections between the latter two parts (Fig. 2). In each SSL system, all these three parts are important for a functional system. In the following paragraphs, the function of each part of the system will be explained [8]. IES standardization in “ANSI/IESNA RP-16-05 Addendum A” [12] has tried to put standard names for different parts of an SSL system, although it does not cover lots of its details.

Reliability in SSL system

A general definition of reliability is; the ability of a product or a system to perform as intended for a specific time in its life cycle environment [20]. The system is an SSL system in this discussion. A system can be classified whether it is functional from the measurable outputs of the system. In this case, the measurable output of a basic SSL system is its output light intensity and color of the light. The SSL system life cycle includes all stages of a life of a product from manufacturing,

Our approach to SSL “Design for Reliability”

Since it has been mentioned in the preceding sections of this paper that SSL is a new technology and there is not enough reliability researches from the literature, there is an essential need for a reliability study of SSL system. As with any other new technology, there is not much information from literature to be used for design improvement or prediction the functionality during its lifetime. Reliability prediction methods [19] can be based on either four references; test data, field data,

Design a SSL device based on “Design for Reliability”

System diagnostics can be implemented in a system by first knowing its major failure cause and its relationship with the lifetime of the device. During the explaining the general approach for Design for Reliability in SSL system (Fig. 3a, Fig. 3b, Fig. 3c, Fig. 3d) the input can be strain, stress, temperature and, etc. In this case study the focus is on the temperature. In HB-LEDs, the relationship between lifetime and junction temperature is known by their manufacturers [24]. Fig. 4 shows an

Conclusions

Providing reliability information and designing a more reliable product is an essential requirement in solid state lighting (SSL) products. However, due to the fact that SSL is a relatively new lighting technology which is expanding the domain of its applications every day, the need for more precise reliability data becomes more critical. The range of SSL applications varies from simple light bulbs for in-door applications to street lighting, automotive lighting and also a wide range of other

Acknowledgements

This research was carried out within the Project “Nanoelectronics for Safe, Fuel Efficient and Environment Friendly Automotive Solutions” (SE2A), co-funded by the European Commission and Agentschap NL in the framework of the Public-Private Partnership ENIAC JU. The hardware implementation of this work is a part of ENSURE project which is the collaboration between Philips Lighting B.V. and TUDelft University of Technology. We also thank people from Philips lighting and Philips research in

References (28)

Held Gilbert
Introduction to LED technology and applications
(2008)
US Department of Energy. Solid-state lighting research and development: manufacturing roadmap; September...
Arturas Zukauskas et al.
Introduction to solid state lighting
(2002)
US Department of Energy (DOE). Solid state lighting manufacturing roadmap, Vancouver, Washington; June...
Richard Karney. Review of the ENERGY STAR® solid state lighting luminaire criteria. Solid-state lighting market...
US Department of Energy. Using LED to its best advantage, LED application series;...
Navigant Consulting, Inc. Multi-year program plan FY’09-FY’15, prepared for US department of energy, Radcliffe...
Patrick Mottier
LED for lighting application
(2009)
Steve Winder
Power supplies for LED lighting
(2008)
Cheng Guo, Guoqi Zhang, Henk van Zeijl, Ignas Niemegeers. Controlling large lighting systems with networked wireless...

Cheng Guo, Henk van Zeijl, Guo Qi Zhang, R. Venkatesha Prasad. An integrated large SSL system with wireless...

ANSI/IESNA RP-16-05 Addendum A. Nomenclature and definitions for illuminating engineering. IES Standard;...

US Department of Energy (DOE). CALiPER program;...

ASSIST Recommendation. LED life for general lighting: definition of life, vol. 1, issue no. 1; February...

Cited by (35)

A Gamma process-based degradation testing of silicone encapsulant used in LED packaging
2021, Polymer Testing
Citation Excerpt :
Light-emitting diodes (LED) are one of the new generation of light sources and are gradually substituting traditional incandescent and fluorescent lamps due to their high luminous efficiency, compact size, high reliability, and environmentally friendly [1–3].
Silicone encapsulant is widely used in light-emitting diode (LED) packaging because it offers high light transmittance, high refractive index, high thermal stability, and long lifetimes. However, it is extremely sensitive to moisture when the LED operates under high temperature and high humidity. In this study, constant moisture stress-accelerated degradation tests (CSADT) are designed with three different thermal stresses on silicone encapsulant, and its real lumen decay and color shift lifetimes are predicted by integrating the accelerated lifetime model with the Gamma process model considering the random effects of the degradation data. The results show that: (1) the lumen decay path after natural logarithmic transformation as well as the color shift degradation path can be well fitted with the linear model. In addition, the degradation rate is related to the thermal stress under the same humidity stress condition; (2) the Gamma process model has a high prediction accuracy when the lumen decay lifetime is used for estimation. On the contrary, the least squares regression (LSR) model has shown superior prediction accuracy versus the Gamma process model when color shift degradation data are used.
Effects of silicone lens aging on degradation kinetics of light-emitting diode package in various accelerated testing
2020, Optical Materials
Citation Excerpt :
The lumen decay amount of LP-CTSS after 2000 h testing is almost equal for SUSADT and SDSADT, although their thermal stresses are loaded in opposite order. Fig. 5 illustrates that the lumen decay paths of LP-CTSS for all individual stresses of SSADT are fitted with the exponential model (Equation [1]). All goodness of fit (R2) values are above 0.89.
Silicone is commonly used in light-emitting diode (LED) package as lens and/or encapsulation owing to its high light transmittance, large refractive index, and good thermal stability. However, the effect of silicone lens aging on the lumen decay and color shift of LED package is a considerable concern for package level reliability. This study empirically investigates the effects of silicone lens aging on lumen decay and color shift of LED package under three accelerated tests, namely, the constant stress-accelerated degradation test, step-up stress-accelerated degradation test, and step-down stress-accelerated degradation test (SDSADT). Results reveal that silicone lens aging substantially contributed to lumen decay and color shift of LEDs and considerably influenced the degradation kinetics of LED package. The color coordinates of LED package covered with aged silicone specimen in three testing methods exhibit a steady yellowing shift law at the same 45-degree direction. Silicone lens aging primarily reduces its transmittance to blue light as accumulated thermal oxidation and hydrolysis reaction, and its effect on the blue peak intensity of LED is independent of the stress-loaded order of step stress-accelerated degradation testing (SSADT). In addition, the effect law of silicone lens aging on LED lumen and color shift could be presented with the exponential and power function models, respectively. Lumen decay and color shift are mostly independent of the historical stress and loaded order in SSADT. The SDSADT testing manner more effectively distinguishes the color shift effect of thermal stress level compared with other accelerated tests. These findings are assumed to benefit the design and reliability assessment of LED products.
Application of expectation maximization and Kalman smoothing for prognosis of lumen maintenance life for light emitting diodes
2018, Microelectronics Reliability
Citation Excerpt :
Light-emitting diodes (LEDs) are popularly used in recent times for indoor lighting, street lamps, communication, medical devices, advertising displays, electronic gadget screens, decorative lighting and much more [1]. LED light sources are known for their high efficiency, environmental resilience, mechanical toughness, high reliability and long lifetime, with claims of the mean time to failure extending to 50,000 h or longer [2–4]. However, owing to their inherent prolonged reliability [5,6], it becomes expensive and time consuming for LED developers to predict the life for LEDs using traditional accelerated run-to-failure life testing, which provide us with the statistical distribution of the time to failure data.
Light emitting diodes (LED) have found widespread use for lighting and displays in recent times due to their high efficiency, favorable form factor, impact resistance, robustness, reliability and prolonged lifetime. The extended times to failure of LEDs make the problem of lumen maintenance life prediction for LED light sources using traditional reliability-based failure data collection and analysis methods quite challenging. Therefore, one has to resort to using prognostic approaches such as Kalman or particle filters applied to real time degradation data to make inferences on the remaining useful life (RUL) of such devices. The standard prognostic approach for predicting lumen maintenance life relies on an underlying degradation model with predetermined parameters from previous stress tests. However, these model parameters may vary quite a bit from device to device and for different manufacturing batch lots, thereby leading to large prediction errors. To address this issue, it would be better to learn the model parameters from the current measurement data directly. This study aims to achieve this by estimating the parameter values using the Expectation Maximization (EM) algorithm. Kalman smoothing is then applied to the identified parameter values for predicting the lumen maintenance life of the LED. The accuracy of the EM – Kalman smoothing approach was tested and compared with the standard nonlinear least square (NLS) approach prescribed by the TM-21 standard as well as the standard particle filter approach (without EM). Our results show that the EM method can give better, if not, similar RUL prediction accuracy with respect to the NLS and standard particle filter (PF) algorithms. Moreover, the accuracy of RUL prediction using our proposed algorithm is insensitive to the choice of the initial model parameter values, which paves way for this algorithm to be used in practice for automated predictive analytics of degrading electromechanical systems.
Step-down accelerated aging test for LED lamps based on nelson models
2017, Optik
In order to quickly acquire the lifetime and reliability prediction for LED lighting products, the method of step-down accelerated aging test based on Nelson models is experimentally studied in this paper. Ten Lide lamps and ten Philips lamps are used in the accelerated aging tests, with humidity less than 50%. The ambient temperatures are 90 °C and 70 °C for Lide lamps, and are 90 °C and 80 °C for Philips lamps. The models, the equal cumulative amount of degradation and the equal cumulative failure probability, are used to obtain the equivalent test time from one stress to another. The maximum likelihood function method is used to obtain two unknown parameters of Weibull distribution at the accelerated ambient temperatures. Then the distribution of the failure probability of LED lamp at ambient temperature of 25 °C is obtained by Arrhenius model. The lifetime tests recommended by Energy Star standard are conducted to verify the accuracy of the models. It is shown that the model of equal cumulative amount of degradation is with a smaller error of lifetime estimation which is respectively 1.74%, 2.41% and 6.57% at the failure probability of 63.2%, 10% and 1% for Lide lamps, and is respectively 8.19%, 1.06% and 5.89% for Philips lamps in this research. The total accelerated aging time about 1000 h is achieved in this experiment, which is quite satisfactory for a quick lifetime prediction of LED lamps recently.
A Critical Review of Design for Reliability - A Bibliometric Analysis and Identification of Research Opportunities
2017, Procedia Manufacturing
It is fundamental to adopt a Design for Reliability (DfR) approach at the stage of new product development in order to guarantee the reliability of an item at all stages of its life cycle. The amount of publications about DfR is not very extensive and it is somewhat dispersed. Thus, the main objective of this article is to collect and perform an analysis of the most recent literature. Fifty publications were analysed with the purpose of identifying the theoretical foundations regarding this topic. Moreover, the main applications, challenges and limitations of DfR were identified aiming the direction for future research.
Thermal degradation kinetics of LED lamps in step-up-stress and step-down-stress accelerated degradation testing
2016, Applied Thermal Engineering
Citation Excerpt :
Light-emitting diodes (LEDs) have been the focus of increasing attention because of their advantages over traditional light sources, such as high luminous efficiency, energy savings, and long lifetime [1].
Step-down-stress accelerated degradation testing (SDSADT) for light-emitting diode (LED) lamps is proposed. Meanwhile, achieving a solution to terminate the initial increase in the optical parameters of LED products rapidly is considered. The thermal degradation kinetics of LED lamps is investigated by SDSADT and step-up-stress accelerated degradation testing (SUSADT) to compare the decay effects of different thermal stress orders loaded in both tests. Two types of commercial LED lamps with different capabilities of heat dissipation (CHDs) are subjected to the investigation. Results show that the inherent initial improvement of the LEDs in SUSADT is more evident than that in SDSADT, and SDSADT can effectively alleviate the initial increase in the optical parameters of LED lamps. The thermal degradation kinetics of LED lamps with strong CHD is only related to the present loaded stress but unrelated to historical loaded stress. This finding is consistent with Nelson’s assumption on step-stress testing. Furthermore, the packaging materials, which include phosphor and silicon layers, play an important role in the degradation of the entire LED package, followed by the degradation of the LED chip. Correlation analysis shows that the degraded optical power of stressed LEDs exhibits a close correlation with the drooped tunneling saturation current.

View all citing articles on Scopus

View full text