Reliability testing and modeling of linear image sensor devices

Electronics packaging is a fast emerging technology. Increase complexity of circuits and reduction in size has aided this growth. It is necessary to build and test products in a smaller time window. Reliability of product is an important and integral part of product quality. Moreover, the designed usable life of the components is long making it impossible to test these products for reliability at normal conditions. Reliability tests need to be performed in a reasonable amount of time and accelerated life testing is used for the same. Accelerated testing is a method of accelerating failures in a product by increasing the amount of stress during the test as compared to normal stress. Normal stress is the one that it would experience during its service life. Failure rate and Mean Life of the product can be determined by collecting failure data on these life tests and interpolating them to useful life.; In this research an effort was made to evaluate the performance of linear image sensor devices as regards to its reliability. Two types of devices; ELIS 1024 thermally cured epoxy and LIS 1024 U-V cured epoxy components; were tested for different reflow conditions (eutectic Sn-Pb and Pb free) and also different epoxy materials used for glass attachment. In order to evaluate the reliability of the components, following accelerated tests were conducted in accordance with the JEDEC standards: (1) Thermal Cycling - JESD22-A104-B; (2) Steady state Temperature Humidity Bias Life Test - EIA/JESD22-A101-B; (3) Temperature Bias and Operating Life - JESD22-A-108-B.; Failure data was collected in each of the tests by taking intermediate readouts and mean life were calculated using chi-square distribution. The failure rates were calculated as summation of failure rates from each test and were estimated values. Failure analysis was carried out to compare failure modes observed in each case.; The test results indicated that there were more failures on the humidity test as compared to HTOL and thermal cycle tests. There were no failures on any of the components types due to temperature variations and it was concluded that the devices are not affected by temperature changes. Some of the observations on the failed devices were bond pad damage, bond pad misregistration, glass misalignment, epoxy color change in case of UV cured epoxy LIS Pb-free parts, fogging of glass, epoxy runout and patches.; A comparison of the calculated MTTF for the three test cells indicated that ELIS Sn-Pb components with thermally cured cover plate epoxy had the best reliability (∼ 5 years) followed by the Pb free ELIS with thermally cured epoxy (∼ 3 years) and LIS Pb free with UV cured epoxy (∼ 3 years).