RU2100817C1 - Method for testing reliability of electronic equipment - Google Patents

Abstract

FIELD: electronic equipment which has extra-long service life. SUBSTANCE: method involves application of electric impact to first group of articles, application of first and second group of articles to ionizing radiation, measurement of electric characteristics of articles in both groups and judging results. Ionizing radiation level is equal to maximal permissible non- destructive dosage, electric impact for first group of articles is equal to operation conditions. Mean value of characteristics of articles in each group is calculated in order to get judgment. EFFECT: increased reliability due to possibility to detect hidden faults caused by electric impact. 1 dwg

Description

 The invention relates to the field of testing electronic products, mainly microelectronics products used in apparatus with long life.

 Known methods for assessing compliance with the requirements for the reliability of electronic products according to GOST B 20.57.404-81 "Products of electronic equipment, quantum electronics and military electrical. Methods of compliance with the requirements for reliability." This standard provides for the identification during testing of catastrophic failures occurring in products due to a violation of their manufacturing technology and design imperfections.

 Testing of products is carried out in the maximum permissible electrical modes of operation and at a temperature maximum permissible for this type of product and the change in the set (criteria) parameters is recorded. The duration of the tests is indicated (set) in the technical specifications for the product. If necessary, tests are also carried out under the influence of other external factors. Products are considered to have passed the test if there were no failures during the test.

 However, the results of these tests are not sufficiently reliable due to the presence of hidden structural defects in the products that cannot be detected by these tests, because they do not affect the change in the criterial parameters of products, and therefore are not amenable to measurement and control. As a result, products recognized as having successfully passed the test may not provide the necessary resource under long-term operating conditions.

 The closest in technical essence and the achieved technical solution is a test method for the reliability of integrated circuits at the stage of manufacture of plates according to US patent N 4816753, M. cl. G 01 R 31/26, publ. 03/28/89. The method is intended to accelerate the testing of products for reliability. For this, the aging of integrated circuits, in particular, MOS integrated circuits, is stimulated by exposure to them with a dose of ionizing radiation (x-rays, etc.) and a correlation dependence of the radiation resistance and the service life of the device is established. As a result of exposure to ionizing radiation in the oxide layers of the device, a large number of electron-donor pairs are stimulated, which are captured by traps at the interface, which leads to failure of the device within a relatively short time.

 Initially, at the calibration stage, one group of products is subjected to electrical aging, i.e., they are exposed to electrical loads until the device fails, which occurs as a result of degradation of traps at the interface and measures the product’s lifetime depending on the substrate current. The measurement of the substrate currents and the lifetime of the products at the stage of their electrical aging is carried out for products manufactured by various technological processes and having different geometric dimensions, in particular different widths. Based on the data obtained by the appropriate mathematical formula, taking into account the width of the device, determine the coefficient of radiation resistance.

 Then, at the test stage itself, the integrated circuits of the other group are exposed to ionizing radiation to simulate the degradation of traps at the interface. The energy of the beam is set sufficient to generate electron-hole pairs induced by ionizing radiation in the oxide layers. The radiation dose is increased to critical, at which the number of traps at the interface leads to the failure of the integrated circuit.

 By comparing the measured critical radiation dose with the radiation resistance coefficient, the product life can be predicted in a very short test time.

 This test method is destructive because the identification of the empirical dependence of the service life of the device on the level of its radiation resistance involves exposure to the plate with a destructive dose of ionizing radiation.

 However, the test results of the known method do not have a sufficiently high reliability. This is explained by the following.

At the stages of electric aging, products receive electrical loads, which, by themselves, regardless of the errors introduced by technological operations in the manufacture of devices, can lead to latent defects in the devices, which during subsequent operation of the products can cause them to fail. However, these data at the calibration stage are laid down as reference data on the radiation resistance of the product. In addition, graduation is performed on some products, while others are tested. As a result, without taking into account possible experimental errors and inevitable differences in the structure of the tested and reference products, the obtained data on resistance are accepted as the source for the analysis of the test results of the studied products. Moreover, as shown by the calculations made according to the description, only due to errors in estimating the current value of the substrate of products at ± 5%, we can expect an error in estimating the life time of products up to 50%
The aim of the invention is to increase the reliability of the tests.

 To achieve the goal of the proposed method of testing the reliability of electronic products, which consists in the fact that the first group of products is first exposed to electrical loads, and then the second and first group of products are exposed to ionizing radiation, measure the specified parameters of the products of both groups, which are judged about the test results. Both product groups are selected from one test batch. Thus, the first group of products is additionally exposed to ionizing radiation. The dose of ionizing radiation for the products of both groups is set to the maximum permissible, non-destructive for this type of product, and the electrical loads for the products of the first group are set within the specified operating conditions, the average statistical value of the specified product parameters in each group is calculated, the results of the tests are compared by comparison.

 The proposed method improves the reliability of the result, because allows you to find out: lead (and how much) or not, traditional reliability tests by exposure to electrical loads on the tested products to the occurrence of latent defects in them, which during subsequent operation of the products can lead to their failure of the previously predicted service life, and if such a fact takes place , then by a comparative assessment of its value to establish how appropriate it is to use these products in the equipment for which they are intended. This is especially important for durable products.

 If the average statistical value of the given product parameters in each group coincides, a conclusion is drawn that there are no effects of the tests themselves on their results. In case the indicated values differ, it is concluded that the test results for reliability are exceeded. In actual use, they will be lower. Depending on how large these differences are, it is possible to recommend the use of products in equipment with more stringent or weaker operational requirements.

 Analysis of the conformity of the claimed technical solution to the criterion of "novelty" showed that no method similar to the claimed was found in the studied and related fields of technology.

 Analysis of the conformity of the claimed technical solution to the criterion of "significant differences" showed that the invention is implemented on the basis of a new set of essential features. Each of the technological operations of the method, taken separately, is necessary, and taken together, taking into account their sequence and interconnections, are sufficient to distinguish the proposed method from other methods and characterize it in the quality that manifests itself in increasing the reliability of the test results.

 The method is illustrated by the drawing, which shows a graph of the relative value of the logical zero voltage from the radiation dose for two groups of products (curves 1,2) from one batch.

 The physics of the processes occurring in the products, in accordance with the proposed test method, can be explained as follows.

 The method is based on the common processes of thermal and radiation aging of materials and products, which is manifested in the identity of the results of the necessary degradation of the products themselves due to the influence of either thermal current loads or ionizing radiation. Formally, this manifests itself in the analogy of the kinetics of parameter drift, for example, for germanium transistors with respect to the parameter, the current transfer coefficient due to long-term functioning (Reliability of Semiconductor Devices, ed. By A.Y. Maslov, Foreign Literature, 1963, p. 284) and under conditions of their irradiation with various types of ionizing irradiations (“The effect of penetrating radiation on electronic equipment”, edited by EA Ladygin, ed. Sov. Radio M. 1980, p. 84). This is due to the commonality of the thermal and radiation mechanisms of the irreversible aging of materials and product structures. The first exponentially depends on temperature, and the probability of each event of material degradation (diffusion, migration, sorption, etc.) is the higher, the lower the activation energy of these processes. For products of semiconductor technology, it ranges from fractions to units of electron-volts (BM Gorin, AE Kiv "Mechanisms of natural aging and forced degradation of semiconductor devices, Reviews on electronic technology", ser. "Semiconductor devices", 1983, no. 8, p. 19, 20 and 31). The second linearly depends on the number of fast particles or quanta and the acts of displacement of atoms of the material caused by them. The probability of each act of material excitation is determined by the activation energy with the material. This energy is an order of magnitude higher than the above.

 As a result, energy is pumped into the field of excitation of the material, which is an order of magnitude greater than necessary for irreversible heat aging processes. Therefore, in these areas all possible irreversible degradation processes begin, similar to thermal processes, whose protective barriers turn out to be much less than the external energy. Thus, ionizing radiation accelerates the usual aging processes of materials and products.

 However, in the proposed method, the action of ionizing radiation not only accelerates the degradation of products, but at the same time, taking into account the additivity principle of accumulation of damage and defects in the product materials, it is possible to identify cases when defects and damage were accumulated a priori. The latter before radiation tests did not appear in any way and could not be detected by methods of measuring the parameters of products or physical analysis of their characteristics.

 When analyzing the results of testing products by comparing the data on the kinetics of changes in the parameters of both groups of irradiated products, those differences in the structure of the studied products that were formed during the testing of products under electric or any other loads are revealed. A more intensive change in the parameters of the studied products during their subsequent radiation tests indicates an irreversible hidden degradation of these products during tests under electric load. The criterion for assessing the degree of this latent degradation is the magnitude of the discrepancy between the kinetics of changes in the parameters of the products of both groups during radiation tests, which allows us to estimate the intensity of the hidden degradation of products under electrical load.

 The inventive method is implemented when testing two batches of bipolar logic circuits of the type 1505LB1A, designed for long-term use.

From the first batch, two groups of products of 50 pieces were selected. each and measured their given electrical parameters, including logical zero voltage
U ^OL .

Then one group was placed in a heat chamber and the durability test mode was set that meets the requirements of TU. After testing, the same electrical parameters were measured. It was found that the parameters, including U ^OL , remained unchanged, which indicated the absence of consumption of the resource of microcircuits, i.e. the test result could be considered true, i.e. reliable.

Then the products of both groups were exposed to a dose of gamma radiation, which was limited to 2 · 10 ⁸ rad, the maximum permissible non-destructive for microcircuits of the type 1505LB1A.

Then we measured the electrical parameters of the products and calculated the average values of the parameters in each of the groups. The relative value U was taken as a criterion parameter $\genfrac{}{}{0ex}{}{\text{OL}}{\text{D}}$ / U ^ol .

U $\genfrac{}{}{0ex}{}{\text{OL}}{\text{D}}$ - the average voltage level after irradiation, B;
U ^{OL is the} average voltage level according to TU, V.

In the group of products tested, the average value of U $\genfrac{}{}{0ex}{}{\text{OL}}{\text{D}}$ / U _OL was equal to 1.5 (curve 1 in the drawing), and in the group of products that did not pass the test, the average statistical value U $\genfrac{}{}{0ex}{}{\text{OL}}{\text{D}}$ / U _OL was 1.2 (curve 2 in the drawing).

 Therefore, the result of previous tests was insufficient, the reliability of the products is lower than previously expected.

Two groups of products were also selected from the second batch of products of the same type and similar operations were carried out with them. In the group of products tested, the value of U $\genfrac{}{}{0ex}{}{\text{OL}}{\text{D}}$ / U _OL was 1.3, and in the group of products that did not pass the test, 1.13.

 Consequently, the result of previous tests was also unreliable. True reliability in this case was lower.

 Based on the test results, we can conclude that in the equipment of long-term functioning it is more advisable to use products from the second batch, since the products of the first batch have a hidden consumption of a parametric resource higher and they may not provide the necessary service life of these products.

 For electronic products of various designs, manufacturing techniques and functional purposes, all standardized parameters of validity criteria, which are informative for electrical tests and radiation exposures, can be measured and accepted for analysis.

 The inventive method really allows to increase the reliability of the tests, since the effect of ionizing radiation is performed simultaneously on the studied and reference (control) groups of products from one batch of their manufacture. As a result of this, the invariance of all functional dependences of the kinetics of measuring product parameters on ionizing radiation and other external disturbances accompanying the tests, for example, temperature, is ensured. Therefore, any test ambiguities during uncontrolled fluctuations of external influences do not affect the reliability of the test results. They are analyzed in the course of comparing data for two groups of products while ensuring complete coincidence of their test conditions. Therefore, in the inventive method, even errors in estimating the absorbed dose of ionizing radiation do not affect the reliability of the analysis of test results of the products.

 Thus, a reliable assessment is made that the products will work reliably, because they have no or minimal parametric resource consumption. In this way, it is possible to evaluate the reliability of products intended for equipment with long-term functioning. This is especially important for devices that are difficult to access.

Claims (1)

 The method of testing the reliability of electronic products, which consists in the fact that the first group of products is subjected to electrical loads, the second group of products is exposed to ionizing radiation, the set parameters of the products of both groups are measured, which are used to judge the test results, characterized in that, in order to increase the reliability of tests, both groups of products are selected from one test batch, the first group of products is additionally exposed to ionizing radiation, and the dose of ion radiation for products of both groups is set to the maximum permissible non-destructive for this type of product, and electrical loads for products of the first group are set within the specified operating conditions, the average statistical value of the specified parameters of the products in each group is calculated, based on the comparison of which the test results are judged.