Method of and program product for performing gate-level diagnosis of failing vectors

• US 20030217315A1
• Filed: 05/12/2003
• Published: 11/20/2003
• Est. Priority Date: 05/14/2002
• Status: Active Grant

• Current Assignee: Siemens Industry Software Inc. (Siemens AG)
• Original Assignee: Logicvision, Inc. (Siemens AG)
• Inventors: Shum, Sonny Ngai San, Maamari, Fadi, Nadeau-Dostie, Benoit
• Granted Patent: US 7,191,374 B2
• Time in Patent Office: Days
• US Class Current: 714/741
• CPC Class Codes: G01R 31/2891 related to sensing or controlling of force, position, temperature G01R31/2874 takes precedence; sensing of force G01L; sensing of position G01B, G01D; sensing of temperature G01K; controlling in general G05
• CPC Class Codes: G01R 31/3177 Testing of logic operation, e.g. by logic analysers

Abstract

A method of fault diagnosis of integrated circuits having failing test vectors with observed fault effects using fault candidate fault-effects obtained by simulation of a set of test vectors, comprises determining a fault candidate diagnostic measure for each fault candidate, the fault candidate diagnostic measure having a fault candidate match metric, an observed fault effect mismatch metric and a fault candidate excitation metric, ranking fault candidates in decreasing diagnostic measure order; and identifying fault candidate(s) having the highest diagnostic measure as the most likely cause of observed fault effects.

Claims

1. A method of fault diagnosis of an integrated circuit having failing test vectors with observed fault effects using fault candidate fault-effects obtained by simulation of a set of test vectors, comprising:

determining a fault candidate diagnostic measure for each fault candidate, said fault candidate diagnostic measure having a fault candidate match metric, an observed fault effect mismatch metric and a fault candidate excitation metric;

ranking fault candidates in decreasing diagnostic measure order; and

identifying fault candidate(s) having the highest diagnostic measure as the most likely cause of observed fault effects.

2. A method as defined in claim 1, said diagnostic measure being a weighted average of said metrics calculated as follows:

(C1*match metric+C2*(1−

mismatch metric)+C3*excitation metric )/3

where C1, C2 and C3 are a match weighting factor, a mismatch weighting factor and a fault candidate excitation weighting factor, respectively.

3. A method as defined in claim 1, said determining a fault candidate diagnostic measure comprising, for each fault candidate:

determining a match metric indicative of fault candidate fault-effects that match observed fault effects;

determining a mismatch metric indicative of observed fault effects that do not match fault candidate fault-effects; and

determining a fault candidate excitation metric indicative of the number failing vectors that have at least one fault candidate fault effect that matches an observed fault effect and of the number of vectors that have at least one fault candidate fault-effect.

4. A method as defined in claim 3, further including:

said determining a match metric including determining a cumulative match metric for said set of test vectors, said cumulative match metric being a ratio of the number of fault candidate fault-effects that match observed faults effects divided by the number of fault candidate fault-effects; and

said determining a mismatch metric including determining a cumulative mismatch metric for said set of test vectors, said cumulative mismatch metric being a ratio of the number of observed faults effects that do not match fault candidate fault-effects divided by the number of observed fault effects.

5. A method as defined in claim 3, further including determining a cumulative match metric for said set of test vectors, said cumulative match metric being a ratio of the number of fault candidate fault-effects that match observed faults effects divided by the number of fault candidate fault-effects.

6. A method as defined in claim 3, said determining a match metric comprising, for each vector of said set of test vectors having at least one fault candidate fault effect, determining the number of fault candidate fault-effects and the number of fault candidate fault-effects that match observed faults effects.

7. A method as defined in claim 6, further including determining a cumulative match metric for said set of test vectors, said cumulative match metric being a ratio of said number of fault candidate fault-effects that match observed faults effects divided by the said number of fault candidate fault-effects.

8. A method as defined in claim 7, further including, excluding from said cumulative match metric, vectors which do not have fault candidate fault-effects.

9. A method as defined in claim 7, further including, excluding vectors from said cumulative match metric which do not have fault candidates fault-effects and observed fault effects.

10. A method as defined in claim 7, further including converting said cumulative match metric to a percentage.

11. A method as defined in claim 3, further including determining a cumulative mismatch metric for said set of vectors, said cumulative mismatch metric being the ratio of the number of observed faults effects that do not match fault candidate fault-effects divided by the number of observed fault effects.

12. A method as defined in claim 3, said determining a mismatch metric comprising, for each vector of said set of test vectors having at least one fault candidate fault effect, determining the number of observed fault-effects and the number of observed faults effects that do not match a fault candidate fault-effect.

13. A method as defined in claim 12, further including determining a cumulative mismatch metric for said set of vectors, said cumulative mismatch metric being a ratio of the number of observed faults effects that do not match fault candidate fault-effects divided by the number of observed fault effects.

14. A method as defined in claim 12, further including, excluding from said set of vectors, vectors which do not have fault candidate fault-effects and observed fault effects.

15. A method as defined in claim 13, further including, converting said cumulative mismatch metric to a percentage.

16. A method as defined in claim 3, said determining a fault candidate excitation metric including, for each vector of said set of test vectors, determining whether said vector includes at least one observed fault effect and at least one fault candidate fault-effect that matches an observed fault effect.

17. A method as defined in claim 16, further including determining a ratio of the number of vectors having at least one fault candidate fault-effect that matches an observed fault effect divided by the total number of vectors having at least one fault candidate fault-effect.

18. A method as defined in claim 5, further including determining a cumulative mismatch metric which is the sum of said number of observed faults effects that do not match fault candidate fault-effects divided by the sum of said number of observed fault effects.

19. A method as defined in claim 18, further including, excluding from said cumulative mismatch metric, numbers associated with vectors which do not have fault candidate faults-effects and observed fault effects.

20. A method as defined in claim 18, further including converting said cumulative mismatch metric to a percentage.

21. A method as defined in claim 18, said determining a fault candidate excitation metric including, for each said vector of said set of vectors, determining whether said vector has at least one observed fault effect and at least one fault candidate fault-effect that matches an observed fault effect.

22. A method as defined in claim 21, said determining a fault candidate excitation metric further including determining a ratio of the number of vectors having at least one fault candidate fault-effect that matches an observed fault effect divided by the number of vectors having at least one fault candidate fault-effect.

23. A method as defined in claim 3, said simulation of test vectors includes simulation of stuck-at faults.

24. A method as defined in claim 3, further including identifying fault candidates as an individual stuck-at-0 or stuck-at-1 fault associated with a circuit node.

25. A method as defined in claim 3, further including identifying fault candidates as a combination of a stuck-at-0 fault and a stuck-at-1 fault associated with a circuit node.

26. A method as defined in claim 3, further including identifying fault candidates as an individual slow-to-rise or a slow-to-fall fault associated with a circuit node.

27. A method as defined in claim 3, further including identifying fault candidates as a combination of a slow-to-rise fault and a slow-to-fall fault associated with a circuit node.

28. A method as defined in claim 3, said overall metric being a weighted average of said metrics calculated as follows:

(C1*match ratio+C2*(1−

mismatch ratio)+C3*excitation ratio )/3

where C1, C2 and C3 are a match weighting factor, a mismatch weighting factor and an fault candidate excitation weighting factor, respectively.

29. A method as defined in claim 28, wherein said weighting factors have equal values C1=C2=C3=1.

30. A method as defined in claim 28, wherein said weighting factors are real numbers between 0 and 1.

31. A method as defined in claim 28, wherein said weighting factors have the following values:

C1=0.4, C2=0.4 and C3=0.2.

32. A method as defined in claim 3, further including,

identifying observed fault effects by applying test vectors to said circuit and comparing circuit output vectors against an expected response;

and

identifying a fault candidate fault-effects by performing a fault simulation of test vectors.

33. A method as defined in claim 3, further including identifying fault candidates comprising:

parsing observed fault effects and identifying all observation points at which a fault effect was observed at least once;

backtracing from each identified observation point to identify cones of logic that reach said marked observe points; and

placing all collapsed stuck-at faults in identified cones of logic into a fault candidate list.

34. A method as defined in claim 3, further including, prior to said identifying fault candidates, loading a plurality of vectors and comparing test response data with expected values to determine failing vectors.

35. A method as defined in claim 3, further including, for each said metric:

counting the data samples used to compute the cumulative metric; and

when a metric data sample count reaches a predetermined minimum sample size (MSS), designating a corresponding fault candidate as eligible for dropping.

36. A method as defined in claim 35, further including, when a fault candidate is designated as eligible for dropping, dropping said fault candidate from further consideration when its overall metric falls below a predetermined dropping threshold.

37. A method as defined in claim 36, further including, dropping a fault candidate from further consideration when its overall metric falls below a predetermined dropping threshold.