A semiconductor chip usually has thousands test parameters in order to guaranteed its quality. Hence, a batch of chips’ test data set include thousands of float data. The primary goal of dealing with this test data is to obtain the fault parameter distribution and judge the chip’s quality. It is a challenge due to the large scale and complex relationship of the test data set. This paper presents a novel method to analyze the test data set by meshing the quality theory and scientific data visualization. First, transfer the test data set to a quality classifier matrix Q: a series of quality region is defined based on quality theory, which is the baseline to classify the test data set into different group and mark them with various number. Second, form a quality-spectrum: define a color rule based on the RGB color model and color the quality classifier matrix Q. Hence chip’s quality distribution could be observed through the quality-spectrum. Furthermore, by analyzing the quality-spectrum, the chip’s quality could be quantitative and fault diagnose has a data basic. One case is included to illustrate appropriateness of the proposed method.
It has been a challenge to quantify the credibility of the accelerated storage model until now. This paper introduces a quantitative measurement named the CMADT (Creditability Metric of Accelerated Degradation Test), which quantifies the credibility of the accelerated aging model based on available data. The relevant criterion data are obtained from the natural storage test. CMADT is a credibility metric obtained by measuring the difference in the metric area between the probability distribution of the accelerated storage model and its criterion data. In addition, the accelerated aging model might include multiple parameters resulting in several single-parameter CMADTs. This paper proposes a method that integrates several single-parameter CMADT metrics into a single metric to assess the overall credibility of the accelerated storage model. Moreover, CMADT is universal for different scales of sample data. The cases addressed in this paper show that CMADT helps designers and decision-makers judge the credibility of the result obtained by the accelerated storage model intuitively and makes it easier to compare various products horizontally.
As the preferred architecture for high-speed and high-resolution analog-to-digital converters (ADC), the accuracy of pipelined ADC is limited mainly by various errors arising from multiple digital-to-analog converters (MDAC). This paper presents a multi-dimensional (M-D) MDAC calibration based on a genetic algorithm (GA) in a 12-bit 750 MS/s pipelined ADC. The proposed M-D MDAC compensation model enables capacitor mismatch and static interstage gain error (IGE) compensation on the chip and prepares for subsequent background calibration based on a pseudo-random number (PN) injection to achieve accurate compensation for dynamic IGE. An M-D coefficient extraction scheme based on GA is also proposed to extract the required compensation coefficients of the foreground calibration, which avoids falling into local traps through MATLAB. The above calibration scheme has been verified in a prototype 12-bit 750 MS/s pipelined ADC. The measurement results show that the signal-to-noise and distortion ratio (SNDR) and spurious-free dynamic range (SFDR) are increased from 49.9 dB/66.7 dB to 59.6 dB/77.5 dB with the proposed calibration at 25 °C. With the help of background calibration at 85 °C, the SNDR and SFDR are improved by 3.4 dB and 8.8 dB, respectively.
Degradation data in practical reliability engineering are often scarce and heterogeneous, originating from multiple sources with varying degrees of uncertainty and conflict. Accordingly, this study proposes a hybrid framework that integrates Dempster–Shafer (D-S) evidence theory with the Wiener process for small-sample reliability assessment using multi-source heterogeneous data. First, a probabilistic non-uniform sampling method regularizes varied data sources and computes basic probability assignments (BPA). Second, a weight synthesis mechanism is constructed, where prior weights derived from prior knowledge are updated by evidence similarity quantified through the Expectation–Width (EW) distance, yielding posterior weights. Quantile sequences from each source are then fused via weighted aggregation to generate a time-series probability box. A Wiener process is subsequently employed to model the probability box (P-box) sequence, enabling interval reliability evaluation. Numerical simulations and a satellite gyroscope case study show that the method effectively fuses multi-source data, provides more comprehensive reliability intervals than single-source approaches, and significantly enhances evaluation robustness under small samples. The framework offers a systematic solution for multi-source data fusion and establishes a novel reliability assessment pathway under uncertainty, with broad applicability to aerospace and precision instrumentation systems.