Determination of tolerance limits for the reliability of semiconductor devices using longitudinal data

Abstract

Design and production of semiconductor devices for the automotive industry are characterized by high reliability requirements, such that the proper functioning of these devices is ensured over the whole specified lifetime. Therefore, manufacturers let their products undergo extensive testing procedures that simulate the tough requirements their products have to withstand. Such tests typically are highly accelerated, to test the behavior of the products over the whole lifetime. In case of drift of electrical parameters, manufacturers then need to find appropriate tolerance limits for their final electrical product tests, such that the proper functioning of their devices over the whole specified lifetime is ensured. In this study, we present a statistical model for the determination of tolerance limits that minimize yield loss. The model considers longitudinal measurements of continuous features, based on censored data from stress tests. The tolerance limits are derived from multivariate distributions where the dependence structure is described by different copulas. Based on extensive numerical testing, we are able to provide insights into the properties of our model for different drift behaviors of the devices.

Spline‐Based Drift Analysis for the Reliability of Semiconductor Devices

“…Also, the necessity of providing probability distributions required for the computation of the desired reliability targets, ruled out the usage of methods for the analysis of variance for repeated measurements, as introduced by Davis. [11] The studies of Hofer et al, [12] Healy et al, [13] Albers et al, [14] and Albers et al [15] are related to the present study. While Al-bers et al [15] and Albers et al [14] present models for the computation of test limits for features that are difficult or costly to measure, Healy et al [13] review and extend existing models for setting test limits while maximizing manufacturers' yield.…”

“…While Al-bers et al [15] and Albers et al [14] present models for the computation of test limits for features that are difficult or costly to measure, Healy et al [13] review and extend existing models for setting test limits while maximizing manufacturers' yield. Hofer et al [12] use multivariate extensions of Archimedean copulas with skew-normally distributed marginals to implicitly describe the time dependencies between the measurements of HTOL stress test. In contrast to their approach, the study at hand directly models the drift behavior of electrical parameters and the computation of the test limits relies on this drift model.…”

“…In that respect, the study at hand extends the approach of Hofer et. al., [12] where the probability for an electrical parameter to drift outside of the specification limits is described only for the predefined measurement points.…”

Research Trends and Tasks of 3.1 Movement in Hwaseong Area, Gyeonggi-do

“…Therefore, proper detection can lead to swift improvements in equipment or process issues. Particularly, faults such as micro-arcing, temporary change, and peripheral point are types that cannot be detected by the conventional Lower Specification Limits (LSL) and Upper Specification Limits (USL) interlocks commonly employed in the current equipment [50]. As a result, this provides an effect of reducing temporal losses.…”