Catastrophic and parametric fault detection by dynamic power supply current test

The increase in importance of mixed signal ICs (MSICs) has led to the need to provide reliable test and fault detection methods for these circuits. The problem of successful testing of MSICs is complex. Many tools and techniques have been proposed to optimise the testability of these circuits, most of them considering separately the analogue and the digital parts of the mixed signal circuits. A different approach to the unification of mixed-mode testing is to analyse the power-supply current through the MSIC. We propose techniques for fault detection, based on power supply dynamic current testing in mixed circuits which take into account catastrophic and parametric faults. The inclusion of parametric faults made it necessary to use multivariate statistics to analyse the power supply current signal. We have developed dynamic current sensors for the digital and analogue parts of the MSIC which depend on of the fault type, catastrophic or parametric. Furthermore, the evaluation methods of the built-in dynamic current sensors output has also been considered through digital processing or by means of a multivariate statistical analysis. The differences between the test methods applied to both fault models, catastrophic and parametric, obliged the adjustment of the dynamic power supply current signal so it can be analysed by the same test equipment. These adjustment may be included in the signal processing measured by the sensors. The application of the dynamic power supply current test methods to some benchmark circuits with injected catastrophic and parametric faults allow us to evaluate the efficiency of these methods.