JP5769413B2 - Electronic component quality evaluation apparatus and method - Google Patents

Description

  The present invention relates to an apparatus and method for evaluating the quality and lifetime of electronic components.

  As an example of an electronic component quality evaluation test, a conventional evaluation test of a semiconductor light emitting device will be described. For example, an evaluation test of a semiconductor light emitting device such as an evaluation of lifetime has been performed based on a standard value of a product. For example, in the case of a semiconductor laser diode, the sample is driven until the drive current value for obtaining the rated light output is 1.2 to 1.5 times the initial value. The driving time is, for example, several thousand hours or more, sometimes tens of thousands of hours. The lifetime of the product was evaluated based on the test results of the sample.

  In the case of a semiconductor laser diode, it is generally considered appropriate to define the time until the drive current value for obtaining the rated light output increases to 1.5 times the initial current value as the lifetime. In the case of a light emitting diode (LED), generally, the time until the light emission intensity from the LED surface reaches 50% of the initial intensity is defined as the lifetime.

  As described above, in the semiconductor light emitting device, the evaluation test is performed based on the standard value of the product. The reason is that the degradation mechanism of the optical semiconductor has not yet been fully elucidated. For example, in the case of a semiconductor integrated circuit, failure modes such as failure due to disconnection of metal wiring are clarified. Therefore, the quality of the semiconductor integrated circuit can be evaluated by, for example, an accelerated deterioration test.

  As main deterioration factors of the optical semiconductor, an increase in crystal defects due to heat and a decrease in luminous efficiency due to the increase are assumed. For this reason, accelerated degradation due to temperature has been proposed for optical semiconductors (see, for example, Patent Document 1). However, for example, the acceleration factor due to temperature, or the failure mode may vary depending on the material constituting the optical semiconductor and the structure of the optical semiconductor. Therefore, as described above, the lifetime of an optical semiconductor is generally evaluated based on the operating current for obtaining a constant light output or the result of measuring the light output at a constant operating current (for example, patents). Reference 1 and Patent Reference 2).

Japanese Patent No. 3060698 Japanese Patent Laid-Open No. 61-155576

  When the conventional evaluation method for an optical semiconductor is specifically executed, for example, a plurality of samples are extracted by sampling from the semiconductor light emitting devices of the same lot, and the extracted samples are driven. Therefore, it takes a considerably long time (for example, several thousand hours or more as described above) before the lifetime of the product belonging to the lot is determined. Further, when the lifetime is found, the lot has already been shipped. Therefore, it is difficult to feed back the evaluation result to the lot.

  Furthermore, according to the conventional evaluation method, in order to improve the reliability of the evaluation result of the lifetime, there is a problem that the number of samples used for the evaluation must be increased.

  Furthermore, the problem resulting from the time required for quality evaluation also occurs. Specifically, it is difficult to comprehensively evaluate stresses that are assumed to be applied to products in the market. In this specification, the “market” is a collective term for various environments in which products after shipment are placed, such as actual use environments. Examples of such stress in the market include disturbances such as humidity, temperature, components contained in air around the product, vibration, and temperature cycle.

As described above, it takes a long time to evaluate the influence on the lifetime of the optical semiconductor. For this reason, it takes a long time just to evaluate the effect of a single type of stress on the life of a product. Therefore, in the conventional evaluation method, it has been difficult to evaluate the influence of various stresses on the product life in terms of time. Since it is not possible to comprehensively evaluate the impact on the product lifetime due to stress assumed in the market, there may be cases where the optical semiconductor fails due to an unexpected factor. Such “unexpected problems” include, for example, a high humidity and high temperature in an atmosphere of corrosive gas (for example, H ₂ S gas), or a case where high temperature and vibration are applied to the product.

  Note that the above-described problem is not only caused in an optical semiconductor, but is a problem that may occur in quality evaluation or life evaluation of electronic components including general semiconductor products.

  In view of the problems as described above, the present invention is a technology that can perform quality evaluation and lifetime evaluation of electronic components in a short time with high efficiency, and can obtain highly reliable and reproducible evaluation results. The purpose is to provide.

An electronic component quality evaluation apparatus according to an aspect of the present invention includes a first state before a predetermined environmental stress is applied to the electronic component, and a second state after the predetermined environmental stress is applied to the electronic component. In both of the states, at least one measurement unit that acquires input / output characteristics of the electronic component, and a plurality of test apparatuses respectively corresponding to a plurality of environmental stresses set in advance as predetermined environmental stresses. The plurality of test apparatuses are configured to apply a plurality of environmental stresses to a plurality of samples of electronic components prepared in advance corresponding to the plurality of environmental stresses, respectively. Based on the input / output characteristics in the first state and the input / output characteristics in the second state obtained from each of the plurality of samples by the at least one measuring unit, the quality evaluation apparatus performs electronic measurement for each environmental stress. further comprising a calculator for calculating the standard SN ratio as an index value for evaluating the impact of the environmental stress on the deterioration of parts. The arithmetic unit averages the input / output characteristics in the first and second states for each sample to obtain an average input / output characteristic, and uses the average input / output characteristics to input / output in the first and second states. The standard SN ratio is calculated based on the input / output characteristics in the first and second states of the plurality of samples after normalizing the characteristics.

An electronic component quality evaluation method according to another aspect of the present invention provides input / output characteristics of a plurality of samples of electronic components prepared in advance corresponding to a plurality of environmental stresses. a first step of obtaining in the first state is a state before each applied, and a second step of applying a plurality of environmental stress on the plurality of samples, a plurality of environmental stress in a plurality of samples each Based on the third step of acquiring the input / output characteristics of the plurality of samples in the applied second state, the input / output characteristics in the first state, and the input / output characteristics in the second state, each environmental stress every, and a fourth step of an index value for evaluating the impact of the environmental stress on the deterioration of the electronic components for calculating a standard SN ratio. In the fourth step, the input / output characteristics in the first and second states are averaged for each sample to obtain an average input / output characteristic, and the average input / output characteristics are used to input the first and second states. The output signal characteristics are normalized, and the standard S / N ratio is calculated based on the input / output characteristics of the plurality of samples after the normalization in the first and second states.

According to the present invention, the quality and lifetime of electronic components can be evaluated in a short time and with high efficiency.
Further, according to the present invention, it is possible to obtain a highly reliable and reproducible evaluation result.

It is the schematic which showed the cross-section of an example of white LED. It is the figure which showed the input-output characteristic of white LED shown in FIG. (A) is the figure which showed the measurement result of the actual input-output characteristic. (B) is a diagram showing the results obtained by normalizing the measured data N ₁ and N ₂ in the data N _0. It is the figure which showed schematic structure of the evaluation apparatus which concerns on the 1st Embodiment of this invention. It is the flowchart which showed the evaluation method of the quality of white LED by the evaluation apparatus shown in FIG. It is the figure which showed an example of the measurement result of the input-output characteristic of a ceramic package sample. It is the figure which showed the time change of the standard S / N ratio. It is the figure which showed the standard S / N ratio in the 5th cycle of a test in a tabular form. It is a radar chart which shows the standard S / N ratio in the 5th cycle of a test. It is the figure which showed an example of the measurement result of the input-output characteristic of a resin package sample. It is the figure which showed the time change of the standard S / N ratio. It is the figure which showed the standard S / N ratio in the 5th cycle of a test in a tabular form. It is a radar chart which shows the standard S / N ratio in the 5th cycle of a test. It is the figure which showed typically the influence on the input-output characteristic of white LED by environmental stress. It is the figure which showed the relationship between the cycle number of an environmental stress test, and luminance change rate (DELTA) (beta). FIG. 16 is a diagram illustrating an example of a lifetime distribution of white LEDs estimated from the relationship between the number of cycles shown in FIG. 15 and the luminance change rate Δβ. 6 is a first flowchart illustrating a life estimation method according to Embodiment 2. 5 is a second flowchart showing a life estimation method according to Embodiment 2.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. The same or corresponding parts are denoted by the same reference numerals, and the description thereof will not be repeated.

  The present invention can be applied without being limited to a specific electronic component. In the following embodiments, an optical semiconductor device is shown as an example of an electronic component.

[Embodiment 1]
The first embodiment will be described by taking a white light emitting diode (LED) used for a backlight of a liquid crystal panel or illumination as an example. In the first embodiment, an apparatus and method for evaluating the quality of an optical semiconductor device will be described.

  FIG. 1 is a schematic diagram illustrating a cross-sectional structure of an example of a white LED. Referring to FIG. 1, LED 1 includes LED element 2, lead frame 3, sealing resin 4, package 5, and metal reflective film 6.

  The LED element 2 is fixed on the lead frame 3. The lead frame 3 is fixed to the package 5. The package 5 is made of, for example, ceramic or resin. The interior of the package 5 is filled with the sealing resin 4. The sealing resin 4 is, for example, a silicone resin containing a fluorescent agent. A metal reflection film 6 is formed on the side surface inside the package 5.

  The LED element 2 emits blue light. The blue light from the LED element 2 excites the phosphor contained in the sealing resin 4. The excited phosphor emits yellow light. By mixing the light emitted from the phosphor and the light emitted from the LED element 2, the LED 1 emits white light to the outside.

  FIG. 2 is a diagram showing the input / output characteristics of the white LED shown in FIG. Referring to FIG. 2, when the white LED has an ideal function, the output (light energy) of LED 1 increases linearly with respect to the increase in input (current) to LED 1. Therefore, the light emission intensity (or luminance) of the LED 1 changes linearly with respect to the current input to the LED 1.

  However, when the environmental stress is applied to the LED, the linearity between the input and the output is not satisfied. As shown in FIG. 2, when an environmental stress is applied to the LED, in a region where the input is large, the input and output may be reduced due to, for example, heat generation of the white LED or saturation of the luminous efficiency of the phosphor. The linearity between is not satisfied.

  In addition, the use of white LEDs gradually deteriorates LED elements, package members, phosphors, and the like. Thereby, the brightness | luminance of white LED falls and an optical characteristic changes.

  In general, the operation time until the luminance maintenance ratio with respect to the initial light emission luminance is reduced to 50% is defined as the lifetime of the white LED. The mechanism of the deterioration of the white LED is not clear, but it is considered that the deterioration of the white LED is affected by, for example, the deterioration of the sealing resin due to heat or the decrease in the reflectance of the metal reflection film due to corrosion.

  The inventor has found that the fluctuation of the input / output characteristics of the white LED can be detected with high sensitivity by an analysis method based on the SN ratio of the dynamic characteristics of quality engineering.

  FIG. 3 is a diagram for explaining an outline of a method for calculating a standard SN ratio of quality engineering. FIG. 3A is a diagram showing a relationship between input values and characteristic values in actual measurement data. The input value shown on the horizontal axis of FIG. 3A is the level value of each control parameter, and the vertical axis shows the characteristic value corresponding to the level value.

The variation of N ₁ and N _{2 with} respect to the average (N ₀ ) of before noise application (N ₁ ) and after noise application (N ₂ ) is clarified. For this purpose, for the purpose of excluding the non-linear component of the line, the measurement data before applying noise (N ₁ ) and the measurement data after applying noise (N ₂ ) are normalized with average (N ₀ ) data.

FIG. 3B shows the result of normalization. By normalization, the linearity and variation of the characteristic value (N ₁ ) before the noise application and the characteristic value (N ₂ ) after the noise application with respect to the average value (N ₀ ) can be evaluated at a time.

  The S / N ratio calculated after normalizing the data is called the standard S / N ratio. Since the influence of the input level value can be excluded by normalization, it is possible to quantitatively evaluate the variation. Furthermore, the linearity can be evaluated simultaneously with the variation by normalization. Thereby, in the evaluation of the white LED, a minute change in the characteristics of the white LED can be detected with high sensitivity.

  In a conventional method for evaluating an optical semiconductor device, the rate of change (for example, luminance maintenance rate) from the initial stage of optical characteristics is measured in a state where a rated current is applied to the optical semiconductor. However, in this method, there is a possibility that the behavior of the change in the optical characteristics does not have high sensitivity with respect to a minute change in the characteristics of the optical semiconductor device.

  In the embodiment of the present invention, attention is particularly paid to the energy conversion efficiency of the optical semiconductor device, and the fluctuation due to the disturbance of the conversion efficiency is set as the evaluation characteristic. Specifically, the fluctuation of the input / output characteristics of the optical semiconductor device is evaluated. The input of the optical semiconductor device is a drive current, and the output of the optical semiconductor device is the light emission intensity or luminance of the optical semiconductor device. According to the first embodiment, in the quality evaluation of the optical semiconductor device, it is possible to realize evaluation with higher sensitivity than the evaluation in the conventional rated state.

The standard SN ratio calculation process is shown below.
First, the total variation _ST is expressed by equation (1).

  Next, the signal magnitude r is expressed by equation (2).

n is the number of noises (environmental stress), and n = 2 in the case of FIG. The line formats L ₁ and L ₂ are represented by the following equation (3).

The L ₁ and L ₂ divided by respectively r / n is the estimated value of the inclination beta ₂ of inclination beta ₁ and N ₂ of N ₁ (see FIG. 3 (B)).

  The standard signal-to-noise ratio (η) is expressed as in equation (9) by the above equations (1) to (3) and the following equations (4) to (8). The standard S / N ratio indicates the strength of input / output (high stability) with respect to the environment.

  In this embodiment, first, environmental stress in the market was analyzed using an FT diagram. Based on the results, the following five types of stress were selected.

(1) Temperature (high temperature)
(2) Temperature (low temperature)
(3) Humidity (4) Mechanical stress (5) Corrosive gas It was confirmed in advance that these environmental stresses act on the degradation of the optical semiconductor device independently of each other.

  In this embodiment, the above five types of environmental stress tests are executed in parallel. Since each stress acts on the degradation of the optical semiconductor device independently of each other, a plurality of environmental stress tests applied to the white LED in the market can be performed in parallel. As a result, the test time can be shortened, and the influence on the functional variation of the white LED due to the stress in the market can be comprehensively evaluated. The evaluation apparatus and evaluation method for that purpose will be described in detail below.

  FIG. 4 is a diagram showing a schematic configuration of the evaluation apparatus according to the first embodiment of the present invention. With reference to FIG. 4, the evaluation apparatus 10 includes chambers 31 to 35 as stress applying apparatuses, a data collection unit 21, and a data calculation unit 22.

  The chambers 31 to 35 are configured to apply the environmental stresses (1) to (5) to the evaluation sample independently of each other. For example, chamber 31 corresponds to stress (1), and chamber 32 corresponds to stress (2). The five chambers and the five types of environmental stresses only need to correspond one-to-one.

  Input / output characteristics are evaluated for the samples (white LEDs) in each of the chambers 31 to 35. For this purpose, each of the chambers 31 to 35 is provided with measuring units 11 to 15 for measuring the input / output characteristics of the sample (that is, the light luminance of the sample to which a current is input). The number of samples in each chamber is not particularly limited, and may be an appropriate number that can obtain a reasonable test result. The number of samples in each chamber may be at least one, but may be three or a plurality such as five. In addition, these numbers are shown as an example, and the number of samples in each chamber is not limited to these numbers.

  The current applied to the sample is swept between a value equal to or less than a threshold current value at which light emission starts and a value greater than the rated current value. By applying a current equal to or higher than the rated current value to the sample, the heat dissipation characteristics including the substrate on which the white LED is mounted can be evaluated. In this embodiment, the maximum value of the current applied to the sample is 1.5 times the rated current.

  The data collection unit 21 collects an input value (current value) to the sample and an output value (for example, luminance) corresponding to the input value for each of the measurement units 11 to 15. At least one measurement unit is sufficient, but by providing a measurement unit corresponding to each chamber, it is possible to simplify the process for measuring the input / output characteristics of the sample and collecting the measurement data.

  The data calculation unit 22 is realized by a personal computer, for example. Based on the data collected by the data collection unit 21, the data calculation unit 22 calculates a standard SN ratio of quality engineering according to the above formulas (1) to (9). The data collection unit 21 and the data calculation unit 22 may be integrated into one device or may be realized by individual devices. Moreover, it is preferable that the data calculating part 22 has a function which memorize | stores the standard S / N ratio obtained by the calculation, and a function which outputs the standard S / N ratio.

  The relationship between the input current of the white LED and the luminance of the white LED reflects the energy conversion efficiency of the white LED. The brightness fluctuates due to environmental stress. In this embodiment, the luminance of the white LED is used as the evaluation characteristic. Thereby, evaluation with higher sensitivity than the evaluation in the rated state as in the past can be realized.

  FIG. 5 is a flowchart showing a method for evaluating the quality of white LEDs by the evaluation apparatus shown in FIG. Referring to FIGS. 4 and 5, in step S <b> 1, the input / output characteristics of the sample before stress application (first state) are measured by measuring units 11 to 15. The first state is, for example, an initial state. In step S2, the data collection unit 21 collects measurement data before applying stress from the measurement units 11-15. In step S3, stress is applied to the sample for a certain period of time. In step S4, the input / output characteristics of the sample after applying the stress (second state) are measured by the measuring units 11 to 15. In step S5, the data collection unit 21 collects measurement data after applying stress from the measurement units 11-15. In step S <b> 6, the standard SN ratio is calculated by the data calculation unit 22 according to the above formulas (1) to (9).

  After step S6, the process may return to step S3. In this case, it is possible to construct an automatic measurement system that repeats the measurement of the input / output characteristics of the sample and the calculation of the standard S / N ratio based on the measurement data every time the environmental stress test is performed for a certain time.

  Next, an evaluation result using the evaluation apparatus shown in FIG. 4 and the evaluation method shown in FIG. 5 will be specifically described. As an evaluation sample, a white LED composed of a ceramic package with a high quality record in the market (average life of about 10 years) was used. Three samples were set in each chamber. That is, 3 × 5 = 15 samples were prepared as a whole.

The conditions of each chamber and the frequency of measurement are shown below:
(1) Chamber 31 (high temperature): 120 ° C. (measured every 100 hr)
(2) Chamber 32 (low temperature): −40 ° C. (measured every 100 hours)
(3) Chamber 33 (humidity): 85 ° C./85% (measured every 100 hr)
(4) Chamber 34 (mechanical stress (heat cycle)): −40 ° C./120° C. (measured every 100 cycles)
(5) Chamber 35 (corrosive gas): H ₂ S gas (concentration 3 ppm), humidity 85% (measured every 100 hr)
FIG. 6 is a diagram illustrating an example of measurement results of input / output characteristics of a ceramic package sample. With reference to FIG. 6, the measurement result of the brightness | luminance with respect to input current is shown about three samples in the chamber 33 which implemented the high temperature, high humidity test (200 hr). The rated current is 100 mA, and the current value is swept from a current below the threshold current to 150 mA (1.5 times the rated current). The standard S / N ratio is obtained from the average value of the initial linear format and the average variation of the linear format after the stress application according to the above formulas (1) to (9). In FIG. 6, two measurement results of three samples substantially overlap each other for the measurement results of “initial” and “after high temperature and high humidity”.

  FIG. 7 is a diagram showing the time change of the standard S / N ratio. With reference to FIG. 7, as a result of evaluating the standard S / N ratio obtained for each cycle (100 hours) of the test, a decrease in S / N ratio was confirmed at the fifth cycle. Therefore, the standard S / N ratio corresponding to all environmental stresses in the fifth cycle is set as the standard S / N ratio representing the quality level of the white LED.

FIG. 8 is a diagram showing the standard S / N ratio in the fifth cycle of the test in a tabular format. FIG. 9 is a radar chart showing the standard S / N ratio in the fifth cycle of the test. The results shown in FIGS. 8 and 9 are created by the data calculation unit 22. 8 and 9, it can be seen that the standard signal-to-noise ratio for the H ₂ S gas is the smallest. That is, it can be seen that H ₂ S gas among the five kinds of environmental stresses has the largest influence on the functional fluctuation of the white LED.

  Thus, by calculating the standard S / N ratio for each environmental stress, it is possible to evaluate the strength of each of the plurality of environmental stresses on the deterioration of the product. Alternatively, by comparing standard SN ratios with respect to the same type of environmental stress between different samples (for example, samples extracted from different lots), the degree of deterioration due to the stress can be compared between samples. The data calculation unit 22 may have a function of evaluating the strength of the influence of each of the plurality of environmental stresses on the deterioration of the product based on the standard SN ratio with respect to each environmental stress.

  Referring to the example of FIG. 9, for example, in the data operation unit 22, the environmental stress with the smallest standard S / N ratio has the greatest influence on the deterioration of the product, and the environmental stress with the largest standard S / N ratio has the largest product. Ordering may be performed for a plurality of environmental stresses so that the degree of influence on deterioration is minimized. Similarly, the data calculation unit 22 may compare the degree of deterioration due to the same type of stress between samples and output the comparison result. By using the standard S / N ratio as an index value for evaluating the quality of the optical semiconductor device, the data calculation unit 22 can perform the quality evaluation as described above.

  The result of comparing the degree of deterioration against stress between different samples will be described. For the white LED of the resin package, the same evaluation as that described above was performed. Since the white LED of the resin package has lower moisture resistance than the ceramic package, the average period until the luminance is reduced to 50% of the initial luminance (that is, the product reaches the end of its life) is about 3 years. That is, the white LED of the resin package has a shorter lifetime than the white LED of the ceramic package.

  FIG. 10 is a diagram showing an example of measurement results of input / output characteristics of a resin package sample. With reference to FIG. 10, the measurement result of the brightness | luminance with respect to input current is shown about three samples in the chamber 33 which implemented the high temperature, high humidity test (200 hr). The measurement conditions are the same as the measurement conditions when the results shown in FIG. 6 are obtained.

  FIG. 11 is a diagram showing a time change of the standard S / N ratio. Referring to FIG. 11, a decrease in the SN ratio was confirmed in the third cycle (300 hr) of the test. However, for comparison with the ceramic package sample, the standard S / N ratio corresponding to all environmental stresses in the fifth cycle was set as the standard S / N ratio representing the quality level of the white LED.

  FIG. 12 is a diagram showing the standard S / N ratio in the fifth cycle of the test in a tabular format. FIG. 13 is a radar chart showing the standard S / N ratio in the fifth cycle of the test. From comparison between FIG. 8 and FIG. 12 (or comparison between FIG. 9 and FIG. 13), it can be seen that the resin package sample is generally less resistant than the ceramic package sample in terms of resistance to environmental stress in the market.

  In the conventional evaluation method, the difference in quality level between the ceramic package and the resin package has been confirmed by a continuous energization test on the sample. However, this method requires, for example, 7000 hours or more. On the other hand, in the evaluation method according to the first embodiment, by using the standard S / N ratio, the difference in quality level between the ceramic package product and the resin package product is clarified in a test time of about 500 hours. Can do.

  As described above, according to the first embodiment, the change from the initial characteristic of the input / output characteristic of the white LED is evaluated using the SN ratio of the dynamic characteristic of quality engineering. Thereby, the characteristic at the time of using white LED in area | regions other than a rated value can be evaluated with high sensitivity.

  Furthermore, according to Embodiment 1, a plurality of environmental stress tests are performed in parallel. Therefore, the degree of deterioration of the element due to each environmental stress can be evaluated in a short time. Since the fluctuation of the product characteristics from the initial stage is evaluated using the S / N ratio of the dynamic characteristics of quality engineering, the quality of the product can be evaluated efficiently in a short time. Furthermore, according to Embodiment 1, a highly reliable and reproducible evaluation result can be obtained.

[Embodiment 2]
In the second embodiment, an apparatus and method for evaluating the lifetime of an optical semiconductor device (more specifically, a white LED similar to that of the first embodiment) will be described. The overall configuration of the evaluation apparatus is the same as that of the apparatus shown in FIG. Hereinafter, the evaluation apparatus according to the second embodiment will be described with reference to FIG. 4 as appropriate.

FIG. 14 is a diagram schematically showing the influence of the environmental stress on the input / output characteristics of the white LED. Referring to FIG. 14, let β ₁ be the slope of a straight line indicating the initial input (current) -output (light energy) characteristics. On the other hand, the slope of the straight line indicating the average input-output characteristic in each environmental stress test is β ₂ . Further, a change rate from the initial slope of the straight line indicating the input-output characteristic is represented by Δβ. Δβ = (β ₁ −β ₂ ) / β ₁ is defined. Δβ represents the rate of change in luminance before and after application of environmental stress.

Referring to FIG. 4, measurement units 11 to 15 measure the input / output characteristics of the sample in the initial state, and input / output the sample every cycle of the environmental stress test (for example, 100 hr as in the first embodiment). Measure characteristics. The data collection unit 21 collects input / output characteristic measurement data. First, the data calculation unit 22 calculates the slope β ₁ of the straight line indicating the input-output characteristics in the initial state and stores the value. For each cycle of the environmental stress test, the data calculation unit 22 calculates the slope β ₂ of the straight line indicating the input-output characteristics based on the measurement data from the data collection unit 21, and further calculates the change rate Δβ. .

  FIG. 15 is a diagram showing the relationship between the number of cycles of the environmental stress test and the luminance change rate Δβ. Referring to FIG. 15, the behavior of the luminance change up to the fifth cycle is measured from the input / output characteristic measurement data. From this measurement result, for each of the five types of environmental stress, the amount of decrease in the luminance change rate when the fifth cycle has elapsed from the initial state is obtained. Based on the amount of decrease in the luminance change rate up to the fifth cycle, the number of cycles when the luminance change rate decreases to 50% (that is, when the white LED reaches the end of its life) can be estimated.

  Furthermore, the acceleration factor in the average environment in each environmental stress test was calculated in advance from a stress model of each environmental stress (for example, Arrhenius model for heat). By multiplying the number of cycles (see Fig. 15) where the change rate Δβ of the slope of input and output is 50% or less (see Fig. 15), under the actual usage environment (in the state where environmental stress is applied in the market) The lifetime of the sample can be estimated.

  FIG. 16 is a diagram showing an example of the lifetime distribution of the white LEDs estimated from the relationship between the number of cycles shown in FIG. 15 and the luminance change rate Δβ. Referring to FIG. 16, the horizontal axis represents the average estimated lifetime of white LEDs, and the vertical axis represents the frequency. The “average estimated lifetime” is an estimated lifetime of five samples, and the five samples were subjected to different environmental stress tests. Using the five samples as one unit, a plurality of average estimated lifetimes respectively corresponding to a plurality of units are calculated. A regression curve is created based on the frequency of the average estimated lifetime. For example, if the regression curve is a normal distribution curve, the time corresponding to the center of the distribution is estimated as the lifetime of the white LED.

  FIG. 17 is a first flowchart illustrating the life estimation method according to the second embodiment. FIG. 18 is a second flowchart illustrating the life estimation method according to the second embodiment. Referring to FIGS. 17 and 18, in step S <b> 11, the input / output characteristics of the sample before stress application (first state) are measured by measuring units 11 to 15. The first state is an initial state. As in the first embodiment, the current applied to the sample ranges from a value less than or equal to the threshold current value at which light emission starts to a value greater than the rated current value (specifically, a value that is 1.5 times the rated current value). Swept between.

In step S <b> 12, the data collection unit 21 collects measurement data before applying stress from the measurement units 11 to 15. In step S13, the slope β ₁ of the output (luminance) of the white LED with respect to the input (current) of the white LED is calculated.

Next, in step S14, stress is applied to the sample for a certain time (for example, 100 hours). One cycle of the test is completed by completing the process of step S14 once. In step S <b> 15, the input / output characteristics of the sample after applying stress (second state) are measured by the measuring units 11 to 15. In step S <b> 16, the data collection unit 21 collects measurement data after applying stress from the measurement units 11 to 15. In step S17, the slope β ₂ of the output (luminance) of the white LED with respect to the input (current) of the white LED is calculated. In step S18, the data calculation unit 22 calculates the rate of change Δβ from the initial stage of the input-output characteristics. The calculated change rate Δβ is stored in the data calculation unit 22.

  Subsequently, the data calculation unit 22 determines whether or not the number of cycles is 5 in step S19. If the number of cycles is less than 5 (NO in step S19), the process returns to step S14, and the stress test is performed again on the sample. Thereby, the luminance change rate Δβ is generated every cycle from 1 cycle to 5 cycles. If the number of cycles has reached 5 (YES in step S19), the process proceeds to step S21 (see FIG. 18).

  In step S21, the data calculation unit 22 estimates the number of cycles where Δβ = 0.5 based on the change in the luminance change rate Δβ from the first cycle to the fifth cycle. For example, the number of cycles at which Δβ = 0.5 is estimated by linear approximation based on the luminance change rate Δβ from the first cycle to the fifth cycle.

  In step S22, the data calculation unit 22 estimates the lifetime of each sample. This lifetime is calculated by multiplying the number of cycles calculated by the process of step S21 by an acceleration coefficient. As described above, a value obtained in advance according to the type of stress test is used as the acceleration coefficient. An acceleration factor is selected according to the type of stress test for each sample.

  In step S23, the data calculation unit 22 creates a regression curve based on the average estimated life time of the samples. In step S24, the data calculation unit 22 estimates the total lifetime based on the regression curve. The lifetime calculated in step S24 is used as an index value for evaluating the quality of the optical semiconductor device.

  As described above, according to the second embodiment, as in the first embodiment, a test capable of detecting a change in product characteristics with high sensitivity is employed. Therefore, as in the first embodiment, a highly reliable and reproducible evaluation result can be obtained.

  Furthermore, according to the second embodiment, the correspondence (acceleration coefficient) between the stress applied in the test and the stress actually applied in the market is obtained in advance. By using the acceleration factor, the lifetime of the product can be evaluated from the result of a short test. Further, according to the second embodiment, since the result of the short-time test is a highly reliable and reproducible evaluation result, the life time evaluated based on the result is also highly reliable and reproducible. Result.

  Further, according to the first and second embodiments, a highly reliable evaluation result can be obtained, and therefore, for example, it is possible to refrain from shipping a lot that may have a short lifetime. As a result, it is possible to extend the life of products distributed in the market.

  Further, according to the first and second embodiments, the quality can be evaluated with a small number of samples and the test time can be shortened, so that the energy consumed for the quality evaluation can be reduced.

  Moreover, according to Embodiment 1 and 2, it becomes possible to feed back a quality evaluation result promptly. This makes it easier to work on improving the manufacturing process. Therefore, for example, the yield can be improved. In addition, the amount of products discarded can be reduced by improving the yield.

  In each embodiment, a white LED is shown as an example of an electronic component. However, the present invention is applicable to quality evaluation of optical semiconductor devices. Therefore, the present invention can also be applied to quality evaluation of laser diodes.

  Further, the present invention is not limited to being applicable only to the quality evaluation of an optical semiconductor device. For example, the present invention can be applied to the evaluation of the quality of electronic components whose quality cannot be evaluated simply by electric characteristics alone. Therefore, the present invention can be applied to, for example, a general semiconductor device. As an example of the semiconductor device other than the optical semiconductor device, for example, a monolithic microwave integrated circuit (MMIC) formed of gallium arsenide (GaAs) can be given.

  The embodiment disclosed this time must be considered as illustrative in all points and not restrictive. The scope of the present invention is defined by the terms of the claims, rather than the description above, and is intended to include any modifications within the scope and meaning equivalent to the terms of the claims.

  2 LED element, 3 lead frame, 4 sealing resin, 5 package, 6 metal reflective film, 10 evaluation device, 11-15 measuring unit, 21 data collecting unit, 22 data calculating unit, 31-35 chamber.

Claims (13)

An electronic component quality evaluation device,
Input / output characteristics of the electronic component both in the first state before the predetermined environmental stress is applied to the electronic component and in the second state after the predetermined environmental stress is applied to the electronic component Obtaining at least one measuring unit;
A plurality of test devices respectively corresponding to a plurality of environmental stresses set in advance as the predetermined environmental stress,
The plurality of test devices are configured to apply the plurality of environmental stresses to a plurality of samples of the electronic component prepared in advance corresponding to the plurality of environmental stresses, respectively.
Based on the input / output characteristics in the first state and the input / output characteristics in the second state obtained from each of the plurality of samples by the at least one measurement unit, for each environmental stress, further comprising a calculator for calculating the standard SN ratio as an index value for evaluating the impact of the environmental stress on the electronic component degradation,
The arithmetic unit averages the input / output characteristics in the first and second states for each sample to obtain an average input / output characteristic, and uses the average input / output characteristics to calculate the first and second states. the input-output characteristic by normalizing, you calculate the standard SN ratio based on the input-output characteristics in the plurality of said first and second states of the samples after normalization, the electronic component quality at Evaluation device.   2. The electronic component quality evaluation device according to claim 1, wherein each of the plurality of environmental stresses is a stress selected in advance as a stress applied to the electronic component in the market.   The electronic component quality evaluation apparatus according to claim 1, wherein the arithmetic unit performs ordering on the degree of influence on deterioration of the electronic component with respect to the plurality of environmental stresses using the standard SN ratio. . The at least one measurement unit is a plurality of measurement units provided corresponding to the plurality of test apparatuses, respectively.
The electronic component quality evaluation apparatus comprises:
The quality of the electronic component according to claim 1, further comprising a data collection unit that collects data of the input / output characteristics from each of the plurality of measurement units and supplies the data to the calculation unit. Evaluation device.   The electronic component quality evaluation apparatus according to claim 1, wherein the electronic component is an optical semiconductor device.   The electronic component quality evaluation apparatus according to claim 5, wherein the plurality of environmental stresses include high temperature stress, low temperature stress, humidity stress, mechanical stress, and corrosive gas stress. The input is a drive current of the optical semiconductor device,
The electronic component according to claim 5, wherein the driving current varies between a value equal to or less than a threshold current of the optical semiconductor device and a value that is 1.5 times a rated current of the optical semiconductor device. Quality evaluation equipment. A method for evaluating the quality of electronic components,
A first state in which input / output characteristics of a plurality of samples of the electronic component prepared in advance corresponding to a plurality of environmental stresses are in a state before the plurality of environmental stresses are applied to the plurality of samples, respectively. A first step of obtaining in
A second step of respectively applying the plurality of environmental stresses to the plurality of samples;
A third step of acquiring input / output characteristics of the plurality of samples in a second state in which the plurality of environmental stresses are respectively applied to the plurality of samples;
An index for evaluating the influence of the environmental stress on the deterioration of the electronic component for each environmental stress based on the input / output characteristics in the first state and the input / output characteristics in the second state and a fourth step of calculating a standard SN ratio as the value,
In the fourth step, an average input / output characteristic is obtained by averaging the input / output characteristics in the first and second states for each sample, and the average input / output characteristics are used to calculate the first and second inputs. the input-output characteristic by normalizing in the state, you calculate the standard SN ratio based on the input-output characteristics in the plurality of said first and second states of the samples after normalization, the electronic component Quality evaluation method.   The quality evaluation method for an electronic component according to claim 8, wherein each of the plurality of environmental stresses is a stress selected in advance as a stress applied to the electronic component in the market. The electronic component quality according to claim 8, further comprising a fifth step of performing ordering on the degree of influence on deterioration of the electronic component for the plurality of environmental stresses using the standard SN ratio. Evaluation method.   The quality evaluation method for an electronic component according to claim 8, wherein the electronic component is an optical semiconductor device.   12. The electronic component quality evaluation method according to claim 11, wherein the plurality of environmental stresses include high temperature stress, low temperature stress, humidity stress, mechanical stress, and corrosive gas stress. The input is a drive current of the optical semiconductor device,
The electronic component according to claim 11, wherein the drive current varies between a value equal to or less than a threshold current of the optical semiconductor device and a value that is 1.5 times the rated current of the optical semiconductor device. Quality evaluation method.

Images