Abstract. We describe an algorithm to build a graphical model—more precisely: a join tree representation of a Markov network—for a steady state analog electrical circuit. This model can be used to do probabilistic diagnosis based on manufacturer supplied information about nominal values of electrical components and their tolerances as well as measurements made on the circuit. Faulty components can be identified by looking for high probabilities for values of characteristic magnitudes that deviate from the nominal values. 1

We describe an algorithm to build a graphical model—more precisely: a join tree representation of a Markov network— for a steady state analog circuit. This model can be used to do probabilistic diagnosis based on manufacturer supplied information about nominal values of electrical components and their tolerances as well as measurements made on the circuit. Faulty components can be identified by looking for high probabilities for values of characteristic magnitudes that deviate considerably from the nominal values. 1.

In this paper, we investigate and systematically evaluate two machine learning algorithms for analog fault detection and isolation: (1) Restricted Coloumb Energy (RCE) Neural Network, and (2) Learning Vector Quantization (LVQ). The RCE and LVQ models excel at recognition and classification types of problems. In order to evaluate the efficacy of the two learning algorithms, we have developed a software tool, termed Virtual Test-Bench (VTB), which generates diagnostic information for analog circuits represented by SPICE descriptions. The RCE and LVQ models render themselves more naturally to on-line monitoring, where measurement data from various sensors is continuously available. The effectiveness of RCE and LVQ is demonstrated on illustrative example circuits. 1 Introduction There are two main uses of testability analysis and fault-diagnosis. Firstly, it results in a set of testability figures of merit (TFOMs) and secondly, it aids in creating a robust diagnostic strategy by identifyi...

Recently, a design-for-test (DFT) methodology for active analog filters was proposed with the primary goal in increased controllability and observability. We operationalize and extend the DFT methodology by using CLP(!) to model and diagnose analog circuits. CLP(!) is a logic programming language with the capability to solve systems of linear equations and inequalities. It is well suited to model parameter tolerances and to diagnose soft faults, i.e., deviations from nominal values. The diagnostic algorithm uses different DFT test modes and voltage measurements at different frequencies to compute a set of suspected components. Ranking of suspected components is based on a measure of (normalized) standard deviations from This is an extended version of the paper that appears in the Working notes Fourth Intl. Workshop on Principles of Diagnosis, DX-93, pp. 105-120, University of Wales, Aberystwyth, September 6-8, 1993. predicted mean values of component parameters. Presented case stu...

The use of cross-correlation between power supply current and output voltage dynamic responses is presented as a methodology for improved mixed current /voltage testing of analogue and mixed-signal circuits. Results obtained from simulations are presented which show that better fault coverage and a simplification of test circuitry can be obtained, when compared with the results obtained from the direct observation of the individual signals. These advantages are important factors to develop viable new design for testability and buil-in self test techniques. 1 Introduction For the manufacturers of electronics products getting a product to market earlier than the competition can lead to higher profits over the product life time. Together with a reduction in throughput time in production, manufacturers are committed to improve products quality and reliability levels. Mixed-signal circuits can be divided into two broad categories which can be identified as Register Control...

We formalize the problem of analog data compression and analyze the existence of a polynomial data compression function. Under relaxed conditions we explore the existence of a solution employing digital signature analysis in the analog domain. 1

Abstract—We consider the problem of identifying a pattern of faults from a set of noisy linear measurements. Unfortunately, maximum a posteriori probability estimation of the fault pattern is computationally intractable. To solve the fault identification problem, we propose a non-parametric belief propagation approach. We show empirically that our belief propagation solver is more accurate than recent state-of-the-art algorithms including interior point methods and semidefinite programming. Our superior performance is explained by the fact that we take into account both the binary nature of the individual faults and the sparsity of the fault pattern arising from their rarity. Index Terms—compressed sensing, fault identification, message passing, non-parametric belief propagation, stochastic approximation. I.

In previous works a technique for doing single fault diagnosis in linear analog circuits was developed. Under certain conditions, one of them being exact circuit parameters, it was shown that only two measurements taken on two selected circuit nodes, at a single frequency, were needed to detect and diagnose any parametric fault. In this paper, the practical value of the technique is evaluated by extending it to the diagnosis of faults in circuits with parameters subject to tolerance. With this in mind, single parametric faults of different strenghts are randomly injected in the circuit under study and, afterwards, these are diagnosed (or the diagnosis fails...). Results are reported on an active filter. Conclusions are drawn on the robustness and effectiveness of the technique.