JEDEC test standards and LED package reliability (MAGAZINE)

May 8, 2012
LED packages that pass JEDEC tests are more reliable in real-life applications, but as yet there is no way to use the results of severe stress-testing of LED packages to project their rated life, as JIANZHONG JIAO explains.
+++++

This article was published in the April/May 2012 issue of LEDs Magazine.

View the Table of Contents and download the PDF file of the complete April/May 2012 issue, or view the E-zine version in your browser.

+++++

With the ongoing development of LED standards in the area of testing LEDs for reliability, the industry is referencing standards that have already been established by long-standing industry bodies such as JEDEC, the Joint Electron Device Engineering Council. For over fifty years, JEDEC has developed testing methods and product standards crucial to the electronics and semiconductor industry, which includes LEDs.

Specifically, JEDEC standards are being relied on heavily in developing standards and regulations for outdoor LED lighting applications such as street and area lighting, where the life requirements of solid-state lighting (SSL) luminaires can extend beyond 50,000 hours and LEDs are exposed to extreme environmental conditions. To ensure that the critical LED packages inside luminaires can maintain a certain lumen output over the claimed life and handle the environmental stress, the LED packages may need to undergo a series of reliability or qualification tests.

The need to evaluate an LED package’s reliability against environmental stress was first introduced in automotive LED lighting applications. The Automotive Electronics Council (AEC) Component Technical Committee has established a list of tests for LED packages. Specifically, the AEC Q101 standard, entitled “Stress Test Qualification for Discrete Semiconductors,” comprises a list of JEDEC standards to be referenced for testing requirements.

In established and/or proposed SSL specifications, JEDEC standards are referred to as part of LED package-level reliability test requirements. This was the case first for the US-DOE-sponsored Commercial Building Energy Alliance (CBEA) Parking Structure Lighting Performance Specification, and later also for the DOE-sponsored Municipal Solid-State Street Lighting Consortium (MSSLC) Model Specification for LED Roadway Luminaires.

Environmental stress

JEDEC tests applicable to LED packages for outdoor lighting applications include three categories of stress to the LEDs. The first category, environmental stress, includes temperature, humidity and salt corrosion and consists of nine separate tests. The LEDs are tested under extreme high and low temperatures while they are in operation, and also in a storage situation. The LEDs are also tested in high temperature and high humidity conditions, in conditions where the temperature is cycled between high and low repeatedly while LEDs are not operating, and also when the LEDs are in operation. These tests are intended to identify the mechanical, electrical and optical degradation of the device induced by temperature and humidity.

There is also a thermal shock test; this is a more severe temperature-related test, where LEDs are transferred from very hot environments (as high as 85°C to 150°C) to very cold environments (as cold as -40°C to -65°C) within twenty seconds. Usually these tests are intended to identify mechanical failures of electrical connections within the LED packages.

Another applicable JEDEC environmental test measures corrosion caused by salt. In both automotive and outdoor lighting environments, salt corrosion of LED components may occur over time, depending on deployed materials. Product quality and a component’s behavior in a corrosive environment can vary greatly from one manufacturer to another, thus testing under these conditions is critical. Corrosion can lead to mechanical failure of electrical connections, as well as optical degradation. In practice, LED manufacturers have identified that other chemicals in the environment, or the outgassing of other components inside the SSL luminaire, can lead to corrosion. This can cause the LEDs to fail or can result in reduced performance. Because the technologies and materials used in LED packages and luminaires vary, the tests for other types of chemical corrosion are usually determined by LED or SSL manufacturers.

Mechanical stress

The second category of LED-package stress – mechanical stress – requires two tests: a mechanical shock test and a variable-frequency vibration test. In automotive or roadway lighting applications, the luminaires are subjected to both mechanical shock and vibrations while being installed and in usage. There are many mechanical interfaces and connections within an LED package that may not survive severe and repeated impact or vibrations over time. These stress tests identify the mechanical behavior or failures of LED packages.

Assembly process stress

The third category assesses LED assembly process stress. This can include two or three tests depending on the LED’s configuration, such as surface-mounted versus through-hole-mounted. The tests are designed to characterize LEDs for moisture or reflow sensitivity and solder heat resistance. LED packages are usually soldered onto a board and then integrated into the final SSL product assembly. During the LED-to-board assembly process, the LEDs can be exposed to extreme temperature. These tests are the basis of the LED package specifications, and LED manufacturers often provide the solder temperature profile in the LED products’ datasheets. It is obviously important that LEDs survive the initial stress of assembly when putting them onto the boards.

Additional JEDEC tests

There are some additional JEDEC standards used in the LED stress tests such as an electrostatic discharge (ESD) test. One test measures ESD sensitivity using the human body model (HBM), identifying an LED package’s susceptibility to damage or degradation by exposure to a defined electrostatic HBM discharge.

The testing or stress conditions used for LEDs are meant to simulate the real-life stresses described above. In practice, in order to minimize testing time while still reflecting real-life stresses, manufacturers use the most accelerated or over-stressed conditions possible without damaging the LED packages. Because JEDEC standards are developed for all solid-state or semiconductor electronic devices, the testing parameters and durations may need to be re-selected to test LEDs specifically.

The JEDEC tests have been widely used by reputable LED manufacturers, initially in LED automotive-lighting applications, and now for outdoor-, street- and area-lighting applications, as the standard qualification tests prior to product launch. LEDs that undergo and pass these JEDEC tests have been proven to be more reliable products in terms of lumen-output maintenance or other performance measures.

Unresolved issues

However, two questions remain. The first is how can the industry standardize the requirements for SSL outdoor lighting in LED package-level qualification, stress or reliability tests? In the above mentioned CBEA and MSSLC specifications, although some JEDEC tests are mentioned or referenced, the detailed requirements are still not clearly specified. For example, the definition of failure criteria is unclear, and test conditions and durations are not standardized. While there was some discussion during the development of the above specifications, the final answer is still unclear.

The second question, which is more challenging to answer, is how to quantitatively link the JEDEC test results of the LED packages to their projected reliability in terms of rated life in number of hours. As mentioned in a previous article, rated life is a statistical measure. If the LED packages passed all of the above JEDEC tests, typically measured in 1000 hours or equivalent time duration, the LEDs were surely exposed to high levels of stress. However, the question remains as to how many hours of real-life operation can be claimed for the LED’s rated life?

The rated life of LEDs is typically stated as 50,000 hours or longer in many SSL outdoor-lighting applications. The link may exist between this rated life and 1000 hours of severe stress-test results, and perhaps this is also a statistical measure. However, there is no standard approach yet established that can identify the link. The industry needs to build a standard procedure, based on the practice of testing LEDs per JEDEC standards and perhaps LM-80 tests, to project the rated life of LED packages.

Currently, it is known that LED packages that pass the above JEDEC tests will be more reliable in real-life applications. In the recent IESNA Testing Procedures Committee (TPC) meetings, a project to pursue standardized LED rated-life projections was proposed, and a working group has formed. Stay tuned for updates on the progress of IESNA TPC standards development.