Rapid evaluation method for major defects of LED indoor lighting products

Because the aging mechanism of traditional lighting source products is relatively clear, such as the consumption of tungsten wire or filament, and the effect of attenuation in a short time is more obvious, the attenuation trend is also close to linear, so the product life can be evaluated in a short time.

However, LEDs are different, and conventional aging is performed at room temperature. In most cases, the light output is still rising in the initial 2000 h, and the lumen maintenance rate at 6000 h is not too obvious, generally maintained at 98%. about. At the same time, according to the industry research results, the light decay curve of the LED chip or module in the 5000h-10000h time period is also inconsistent with the beam curve in the 10000h-15000h time period, which gives the life prediction band through the conventional aging. It is a great difficulty.

That is, the life expectancy of the LED is measured by conventional aging, and it takes at least 15,000 hours to obtain relatively accurate test data.

By applying high temperature stress, the aging speed of LED products can be accelerated, and the life evaluation time of LEDs can be shortened. Taking into account the single failure mechanism characteristics of LED products, the current test methods for accelerated aging evaluation life are mainly for LED chips or LED modules, and do not contain other electronic components. However, the accelerated test was carried out at a shell temperature of 55 ° C and 85 ° C. Under 6000 h, the light attenuation of the LED was not significant. In most cases, the lumen maintenance rate was only reduced to about 93%, and 55 ° C and 85 ° The mathematical correlation of °C data for life evaluation is not too reliable.

Therefore, by studying the rapid assessment method for major defects, the LED product is regarded as an overall system, and it is not evaluated how reliable it is, but how much reliability is evaluated. The rapid assessment method for major defects can evaluate and screen the quality reliability of LED products in a short time (1000h), which is relatively suitable for the development and demand of the semiconductor lighting industry.

2. Introduction of major defects in LED lighting products

Like humans and animals, the main reason for the short lifespan of LED lighting products is that they have "big illnesses", that is, there are "significant defects." These potential major defects, although in the initial stage of the use of LED lighting products, do not affect their normal work, but after a period of time, the defects are excited, which in turn causes the whole lamp to fail. In general, such failures will be triggered within 2-6 months.

The major defects in LED lighting products can be classified into two categories as a whole:

(1) Design defects

The most common cause of such defects is the poor design, the insufficient margin reserved during design, and the inability to support LED lamps for a long time. Life-related design defects, including the output specifications of the control device and the specifications of the light source module, the design of the driver IC, the selection of components, circuit protection, and heat dissipation design.

(2) Manufacturing defects

Manufacturing defects, also known as process defects, are caused by process defects in the manufacturing process, such as solder joint defects, IP protective defects, and the like. These defects are not related to the design, but are caused by the negligence of the workers or the production line during the production process.

Under normal circumstances, design defects and manufacturing defects can only be reduced as much as possible, but not completely eliminated. The existence of design defects and manufacturing defects is the root cause of the high early failure rate of LED lighting products. In the electronics industry, the “bathtub curve [1]”, which describes the variation of product failure rate over the life cycle, has been widely recognized.

The reliability of electronic products has changed from the time of use to the end of life. Taking the failure rate of the product as the reliability characteristic value of the product, the use time is the abscissa, and the failure rate is the ordinate. The variation curve of the failure rate with the use time during the life span is similar to the bathtub, so it is called “bathtub curve”, see Figure 1.

The bathtub curve is divided into three stages, namely the early failure period, the occasional failure period, and the wear failure period. It can be seen from the figure that the failure rate is high during the early failure period, but the failure rate decreases rapidly with the increase of the working time of the product. Most of the reasons for the failure at this stage are due to defects in the design, raw materials or manufacturing process. Caused. The occasional failure period is also called the random failure period. In this stage, the product failure rate is low and stable, and can often be regarded as a constant. The product uses the value of this stage when calculating the MTBF and the product failure rate. During this period, the product performance is stable, and the cause of accidental failure is mainly caused by quality defects, material weaknesses or environmental and improper use. In the period of wear and tear, the failure rate of the product increases rapidly with time, which is mainly caused by product wear, fatigue, aging and material consumption.

The three stages of the above bath tub curve are suitable for most areas, including LED lighting. The information obtained from a large amount of engineering data, the lighting project in the early stage of lighting, about 2-6 months, is often the high incidence of LED lamp failure rate, after this period of time, the situation of bad lights is significantly reduced.

Relative to the failure of the accidental failure period, the failure of the early failure period (hereinafter referred to as early failure) is generally exposed in a short period of time, and the problem is easier to find and the cost of the solution is lower. Summarizing and summarizing the major defects of LED lamps is conducive to discovering the causes of early failures, avoiding the irrational design and eliminating manufacturing process defects, minimizing the early failure rate of products, and thus improving the overall reliability level of batch samples. .

3. Rapid evaluation method for major defects of LED lighting products

3.1 Ideas for rapid assessment of major defects

The rapid evaluation test method for major defects is based on the failure data investigation of large-volume LED products, and analyzes the environmental conditions of the LED products and the causes of failures. In order to be able to excite the defects of the LED lamp in a short time, the test stress should be increased, but the test stress should not exceed the working limit of the product.

Test methods include rapid temperature cycling (or temperature shock), random vibration, alternating heat and humidity, and early reliability assurance.

The components and materials of most electronic products are temperature-sensitive devices, and temperature changes cause thermal expansion and contraction. The thermal expansion and contraction coefficients of various parts and various components of LED lamps are different. When the temperature changes rapidly, it is easy to cause stretching and cracking of electrical connection parts, causing potential cracks and even disconnection of electrical connection layer. The harsher the temperature of the upper and lower limits and the faster the temperature change rate, the more obvious the test effect.

After the rapid temperature change (or temperature shock) test, the random vibration test is performed immediately, and the potential electrical connection defects are further excited. The components with unreasonable selection of the partial volume on the power supply board may even damage the fragile mechanical connectivity. Causes the soldering foot to fall off or break. For some unreliable welding methods, such as virtual welding, lap welding, the test can also stimulate its defects.

After the vibration test, an alternating damp heat test was applied. After the first two steps of testing, it may be that some of the internal components of the LED lamp and the soldering legs of the board may have microcracks, but the surface shows that it can still work normally. At this time, the alternating heat and humidity test is carried out, and the temperature and humidity alternately change. When the temperature changes, a certain "breathing effect" is generated in the lamp body, which exacerbates the moisture entering the lamp body, adsorbing on the components and the soldering feet, and along the The propagation of the welding cracks eventually leads to short-circuiting of the components and the failure of the LED lamps.

In the previous three-step test, there is still no failed sample, and the manufacturing process has a good level. If there is any defect in the design, it will pass the early reliability guarantee test. The early reliability test conditions are subject to high temperature and high humidity conditions, and the test sample is simultaneously applied with switching. The purpose of this part of the test is to fully consider the durability of LED lamps from the system to the device level. The setting of high temperature conditions does not exceed the design limit of the device. Although it cannot make all kinds of components such as capacitor parts and semiconductor devices reach their rated life level effectively, in the accelerated stress test period, the element with qualified quality level The device can maintain normal operation and is still guaranteed.

3.2 Rapid assessment of major defects

Test methods include rapid temperature cycling, random vibration, alternating heat and humidity, and early reliability assurance.

3.2.1 Selection of test samples

The samples tested must have batch characteristics that represent the overall level of the mother. If the test sample is aged and the early failure is removed, the test results are not representative.

Samples for reliability test evaluation should be randomly selected from samples of at least one batch that have undergone a preliminary screening step. The LTPD value specified in the evaluation test is generally 10 or 20, and the corresponding sample number is 22 or 11. When the LTPD value is equal to 10, if one unqualified luminaire is found in the first 22 samples, the next 16 samples should be taken for testing. If there are no unqualified luminaires in the second sample taken, it can be considered that one of the 38 samples in the product failed, passing the LTPD standard of 10%. Similarly, when the LTPD value is 20, if one unqualified luminaire is found in the first 11 samples, the next 7 samples should be taken for testing. If additional unqualified luminaires are found, they meet the LTPD standard of 20%. The sampling plan is shown in Table 1:

Table 1 LTPD sampling plan

3.2.2 Failure criteria for the test

Failure criteria for rapid assessment of major defects, including the following types:

a) LED whole light or several light sources are not bright;

b) The overall light decay of the LED exceeds 6% (this value is the proportion of light decay when the LED chip operates 6000 hours according to the exponential distribution);

c) flashing of the LED whole lamp;

d) the phosphor layer appears to have a falling off of the phosphor layer;

e) The LED color lamp or several light sources have obvious color temperature changes, that is, the apparent illuminating color changes suddenly, such as from "white light" to "blue light";

Because the test scheme adopts the high stress accelerated test evaluation method, the phenomenon of softening deformation of the plastic casing or the like appearing in the test sample is generally caused by the heat resistance temperature of the material being lower than the test temperature. Therefore, such phenomenon is not recommended. The basis for the failure determination.

3.2.3 Determination of test stress and conditions

(1) Rapid temperature cycle test or temperature shock test [2]

The upper and lower limits of the rapid temperature change test or temperature shock test shall be determined by the limit sensitive stress test. Through high-temperature stepping and low-temperature stepping test, the high and low temperature limits TU and TL are determined, and the previous temperature of the limit temperature is used as the basis for the high temperature and low temperature of the rapid temperature cycle test or the temperature impact test.

During the test, the sample was not charged and the function was checked before and after the test. For LED application lighting products, considering the common internal component types and specifications, the recommended test conditions are high temperature of +85 ° C, low temperature of -40 ° C, and holding time at extreme temperature, recommended for 1 h or 2 h, number of cycles For 10 times, the rate of temperature change is at least 10 ° C / min.

The test conditions for random vibration shall be determined according to the purpose and needs of the test.

For tests with high reliability requirements and higher timeliness requirements, the choice of conditional stresses needs to be more stringent. Similar to the temperature step test, before the vibration test, the vibration step stress test is used to find the working limit, and the first-order stress before the vibration working limit is used as the basis for the vibration condition.

For tests with high reliability requirements and intended to stimulate potential defects, it is recommended to perform random vibration tests in accordance with uniform test conditions. The test conditions are as follows:

1) Working conditions: normal working condition

2) Frequency range: 5Hz ~ 200Hz

3) Acceleration spectral density: 0.01g2/Hz

4) Total rms value of acceleration: 1.4 grms

5) Test direction: X, Y, Z three directions (not considering the possible structural symmetry of the actual product, that is, any form of product must be tested in three mutually perpendicular directions)

6) Test time: 15 minutes in each direction.

(3) Alternating damp heat test

Before the setting of the alternating heat and humidity test conditions, the working environment temperature and humidity that the test sample may be in the actual work shall be investigated, and based on the environmental conditions, the test conditions may be improved, for example, the temperature is increased by 10 ° C. Increase humidity by 20% RH.

Several switching on and off operations are applied during the high temperature and low temperature time of each temperature cycle. For general LED indoor lighting products, the recommended test conditions are:

1) Working conditions: normal working state, each cycle in the high temperature and high humidity stage to conduct a sample power on and off, the power off time is not less than 3min, not higher than 6min.

2) Upper limit temperature 55 °C, low temperature sub-circulation: -10 °C

3) Number of cycles: 10 times

The test profile and test curve are shown in Figure 5:

(4) Early reliability assurance test

For the condition setting of the early reliability guarantee test, refer to the GJB 899A and GJB 450A standards. It is recommended to use the “85°C/85% RH168h” test condition. Before the test, the temperature protection circuit on the circuit board must be removed or bypassed. In the test sample, the test samples are divided into two groups, one set is applied to switch on and off, and one set is continuously operated.

If the rated resistance of the components used is low, or the temperature rise of the sample is too high, as an alternative, the "70 °C / 85% RH264h" test condition can be selected. The test conditions are not recommended to be too low, lower temperatures, and take longer to effectively provoke product defects.

Among the LED lighting products, the most commonly used temperature sensitive devices include electrolytic capacitors, semiconductor devices, etc. The heat-resistant extreme temperatures of these two types of devices are generally 125 ° C and 155 ° C. Under normal working conditions, the temperature rise of electrolytic capacitors and semiconductor devices is generally around 20K and 50K, that is, the ambient temperature is 85 °C. The maximum temperature of the two types of device housing is about 105 ° C and 135 ° C, and the distance limit temperature is retained. A certain amount of money.

Of course, the limit temperature of some components is also really low, which is required to adjust the corresponding test temperature according to the actual situation.

4. Types of defects and their effectiveness triggered by rapid assessment methods for major defects

4.1 LED lighting products in the actual use of the failure mode type investigation

There are many reasons for the failure of LED lamps. According to the failure parts, there are two types of light source failure and power failure. According to the research situation, in the actual application, the failure modes of LED lights mainly have the following modes:

4.1.1 Common failure modes of light sources:

(1) Degradation of the ohmic contact of the chip, eventually causing the contact to gradually crack or even fall off;

(2) The aging of the material causes the heat dissipation effect to be deteriorated, and the junction temperature of the LED is gradually increased to accelerate the failure;

(3) The welding process is poor, so that the solder joints are easy to fall off;

(4) electrostatic damage;

(5) LED vulcanization;

(6) The chip packaging process is poor, and it is prone to failure such as gold wire breakage and phosphor falling off;

(7) Electrical breakdown.

4.1.2 Common Failure Modes of Power Supply:

(1) lightning strike failure;

(2) The grid fluctuation causes the power rectifier component to accelerate aging failure;

(3) the solder joints fall off;

(4) aging failure of components, such as electrolytic capacitors, rectifier bridges, transformers, etc.;

(5) The short circuit of the component leads to the circuit burning;

(6) The matching between the power supply and the light source module is poor, so that the power supply efficiency is low when the power supply is working, the internal temperature of the power supply rises, the PFC circuit is prone to failure, and the MOS tube is easy to fail.

4.2 Types of defects triggered by rapid assessment methods for major defects

The samples of large-volume LED lamps were tested according to the rapid evaluation test method for major defects, and the test was conducted using the "22 samples / 0 failure" criterion of the LPD 10% test protocol. The conditions for rapid temperature change, random vibration, alternating heat and humidity and early reliability assurance tests were carried out under the recommended test conditions above. The disassembly and analysis of the failed samples during the test are summarized as follows:

(1) the solder joints fall off;

(2) The short circuit of the components under the humidity test conditions causes the circuit board to burn out;

(3) Fuse, transformer, rectifier bridge, electrolytic capacitor, MOS tube and other components are invalid;

(4) Light bulb breakdown;

(5) The inside of the lamp bead is open;

(6) Material deformation, etc.

From the comparison and analysis of the two parts, it can be seen that the failure mode type of the LED lighting products evaluated under the test conditions of the major defects rapid evaluation method has a high consistency with the failure mode under the normal working conditions, and many normal conditions. The failure conditions that occur underneath can be stimulated and discovered through rapid assessment of major defects. Of course, some types of failures cannot be stimulated by these tests, such as LED vulcanization, electrostatic breakdown, grid fluctuation acceleration damage, etc., but other types of failure can pass the directional reliability evaluation test in a short time. Inspired within.

4.3 Rapid assessment method for major defects Screening ratio of LED indoor lamps

Using the LPD 10% "22 samples / 0 failure" test protocol, the conditions of rapid temperature change, random vibration, alternating heat and humidity and early reliability assurance test adopt the above recommended test conditions for four types of LED indoor lighting products. The test results were counted, including LED downlights , LED spotlights (including MR16 and PAR lights), LED tubes , and LED bulbs . According to the statistical results, the failure rate analysis is shown in Table 2:

Table 2 Statistics of the failure rate of LED indoor lighting

The above test results can be reflected. For LED lighting products without high IP protection design, the various test stresses in the primary test have a reasonable screening ratio. For different types of products, the rate of failure is also different, but the overall failure rate of LED indoor products is also 46.68%.

4.4 Rapid assessment of major defects The life performance of samples that have not failed after the test

Part of the sample will be randomly selected after passing the above test, and the light will continue to work and monitor the life under normal temperature conditions. The continuous lighting time has exceeded 7000h, and the failure rate of each type of LED lamp is shown in Table 3. :

Table 3 7000h failure statistics after continuous evaluation of major defects in LED indoor lights

It can be seen from the above table that the sample after the major defect evaluation test has a very low failure rate during the continuous illumination working period. Equivalent to, after eliminating the major defects, the life can be higher, at least 95% of the reliability of the 7000h life is guaranteed.

5. Differences and correlations between rapid assessment methods for major defects and LED life assessment methods

At present, the current life standards or reliability of LED lamps in national standards and IEC standards are mainly focused on predicting the lifetime by continuously monitoring the lumen maintenance rate for thousands of hours or more, and proposes the lifetime value of L70 or L50. That is, the time when the lumen maintenance rate is reduced to 70% or 50% is the product life. This idea is mainly quoted from the US Energy Star, referring to the evaluation method of traditional lighting sources. However, the difference is that traditional lighting products have a relatively short life span. When they are ignited for 5000h, they are close to the end of life, that is, they are close to the wear and tear period on the timeline of the “bathtub curve”, so this method is used to evaluate traditional lighting. The product has high reliability and relatively economical time. However, the LED has a long life and is evaluated according to this idea. More is to reflect the life of the individual, but it does not represent the overall reliability performance.

Therefore, the accelerated life evaluation test method has become a hot spot in the research of LED lighting industry at home and abroad. At the same time, this is also an international problem that plagues the entire semiconductor lighting industry. The life of an LED application depends on the short board of the lamp system. This short board may be a component, or it may be a material or even a solder joint.

At present, it is impossible to confirm the long life of the LED lamp, but through the major defect assessment test, we can confirm the short life of the test sample. At the same time of proposing this scheme, according to the data of later tracking, when applying the major defect assessment test plan, we can also propose the guaranteed life value with a certain level of reliability. For example, a batch of samples tested can reach a life of at least 7000 hours, which has a 95% reliability level.