Fault dictionary-based scan chain failure diagnosis

• US 8,615,695 B2
• Filed: 06/13/2007
• Issued: 12/24/2013
• Est. Priority Date: 04/04/2007
• Status: Active Grant

• Current Assignee: Mentor Graphics Corporation (Siemens AG)
• Original Assignee: Mentor Graphics Corporation (Siemens AG)
• Inventors: Guo, Ruifeng, Huang, Yu, Cheng, Wu-Tung
• Primary Examiner(s): MCMAHON, DANIEL F
• Application Number:
  US11/818,440
• Publication Number:
  US 20080250284A1
• Time in Patent Office: 2,386 Days
• Field of Search: 714/724, 714/726, 714/727, 714/729, 714/733, 714/738
• US Class Current: 714/737
• CPC Class Codes: G01R 31/3177 Testing of logic operation, e.g. by logic analysers
• CPC Class Codes: G01R 31/318525 Test of flip-flops or latches
• CPC Class Codes: G01R 31/318541 Scan latches or cell details

Abstract

A dictionary-based scan chain fault detector includes a dictionary with fault signatures computed for scan cells in the scan chain. Entries in the fault dictionary are compared with failures in the failure log to identify a faulty scan cell. In one embodiment a single fault in a scan chain is identified. In another embodiment, a last fault and a first fault in a scan chain are identified.

Claims

1. A method of identifying faulty scan cells in a scan chain, comprising:

receiving a failure log that indicates failures detected during scan chain testing of an integrated circuit;

reading one or more fault signatures from a fault dictionary that specifies failures that will occur in a scan chain readout if a scan cell in the scan chain has a fault, wherein the fault dictionary includes a complete fault signature for a first candidate scan cell in the scan chain and an incomplete differential fault signature for a second candidate scan cell in the scan chain from which a complete fault signature for the second candidate scan cell can be recreated; and

with a computer, using the one or more fault signatures stored in the fault dictionary and the failure log to identify a faulty scan cell in the scan chain.

2. The method of claim 1, wherein for at least one scan cell, the fault dictionary stores a list of failures that will occur in a scan chain readout if the at least one scan cell has a fault, and wherein the faulty scan cell is identified by finding a match between the list of failures for the at least one scan cell and the failures included in the failure log.

3. The method of claim 1, wherein the faulty scan cell is identified as the second candidate scan cell by constructing a complete fault signature for the second candidate scan cell using the differential fault signature and finding a match between the complete fault signature and the failures included in the failure log.

4. The method of claim 1,

wherein the using the fault signatures stored in the fault dictionary and the failure log to identify a faulty scan cell in the scan chain comprises determining relative fault signature differences between the differential fault signature stored for the second scan cell and the complete fault signature stored for the first scan cell.

5. The method of claim 1, further comprising storing an identity of the faulty scan cell.

6. A method of identifying a last fault in a faulty scan chain having multiple faults, the method comprising:

receiving a failure log indicating failures detected during scan chain testing of an integrated circuit, the failure log indicating failures on the faulty scan chain and failures on scan chains that are free of faults;

reading a fault signature from a fault dictionary that specifies failures that will occur in scan chains of the integrated circuit when a candidate scan cell in the faulty scan chain has a fault, the failures specified including failures in the faulty scan chain and failures on scan chains that are free of faults the fault dictionary including one or more complete fault signatures and one or more incomplete differential fault signatures; and

using a computer, comparing the failures specified in the fault signature for scan chains that are free of faults with failures in the failure log for scan chains that are free of faults to identify the candidate scan cell as the last fault in the faulty scan chain.

7. The method of claim 6, further comprising storing an identity of the candidate scan cell identified as the last fault.

8. The method of claim 6, further comprising deriving the fault signature from at least one of the complete fault signatures and one of the differential fault signatures.

9. The method of claim 6, wherein the fault is a stuck-at 0 or stuck-at 1 fault.

10. The method of claim 6, wherein the fault is a timing fault.

11. The method of claim 6, wherein the failure log further indicates failures on primary outputs, wherein the fault signature further specifies failures that will occur at primary outputs when a candidate scan cell in the faulty scan chain has a fault, and wherein the act of comparing includes comparing the failures specified in the fault signature for the primary outputs with failures in the failure log for primary outputs to identify the candidate scan cell as the last fault of the multiple faulty scan cells in the faulty scan chain.

12. The method of claim 6, wherein the candidate scan cell is a last fault candidate scan cell, the method further comprising:

reading a fault signature for a first fault candidate scan cell;

creating a composite fault signature for the first fault candidate scan cell, the composite fault signature including failures defined in the fault signature for the first fault candidate scan cell that are upstream of the first fault candidate scan cell and failures defined in the fault signature for the last fault scan cell candidate that are downstream of the first fault candidate scan cell; and

comparing the failures in the composite fault signature with failures listed in the failure log to identify the first fault candidate scan cell as the first fault in the faulty scan chain.

13. The method of claim 6, further comprising:

determining a first fault scan cell by;

comparing a composite fault signature for a candidate scan cell with failures listed in the failure log to identify the first fault scan cell in the faulty scan chain, wherein the composite fault signature for the candidate scan cell includes the fault signature for the candidate scan cell and its upstream scan cells specifying failures that will be detected in a defect-free scan chain as determined from a previously determined last fault scan cell and a fault signature of scan cells that are downstream of the candidate scan cell that will be detected in the scan chain that includes the first fault candidate scan cell.

14. A method of identifying a first fault in a faulty scan chain, the method comprising:

reading a fault signature for a candidate scan cell that may be the first fault in the faulty scan chain;

reading a fault signature for a previously determined last fault in the faulty scan chain, the fault signature being a differential fault signature from which a complete fault signature can be determined with reference to a base fault signature;

using a computer, creating a composite fault signature for the candidate scan cell that includes failures defined in the fault signature for the candidate scan cell that are upstream of the candidate scan cell and failures defined in the fault signature for the last fault that are downstream of the candidate scan cell; and

using the computer, comparing the failures in the composite fault signature with failures listed in a failure log to identify the candidate scan cell as the first fault in the faulty scan chain.

15. The method of claim 14, further comprising storing an identity of the candidate scan cell identified as the first fault.

16. The method of claim 14, wherein at least one of the fault signatures for the candidate scan cell or the fault signature for the previously determined last fault is a relative signature, the method further comprising deriving the at least one of the fault signatures from at least a differential signature and a complete base signature.

17. The method of claim 14, wherein the fault is a stuck-at 0 or stuck-at 1 fault.

18. The method of claim 14, wherein the fault is a timing fault.

19. One or more non-transitory computer readable media storing a series of instructions that are executable by a computer to perform a method, the method comprising:

receiving a failure log that indicates failures detected during scan chain testing of an integrated circuit;

reading one or more fault signatures from a fault dictionary that specifies failures that will occur in a scan chain readout if a scan cell in the scan chain has a fault, wherein the fault dictionary includes a complete fault signature for a first candidate scan cell in the scan chain and an incomplete differential fault signature for a second candidate scan cell in the scan chain from which a complete fault signature for the second candidate scan cell can be recreated; and

using the one or more fault signatures stored in the fault dictionary and the failure log to identify a faulty scan cell in the scan chain.

20. The one or more non-transitory computer readable media of claim 19, wherein for at least one scan cell, the fault dictionary stores a list of failures that will occur in a scan chain readout if the at least one scan cell has a fault, and wherein the faulty scan cell is identified by finding a match between the list of failures for the at least one scan cell and the failures included in the failure log.

21. The one or more non-transitory computer readable media of claim 19, wherein the faulty scan cell is identified as the second candidate scan cell by constructing a list of failures for the second candidate scan cell using the differential fault signature and finding a match between the list of failures and the failures included in the failure log.

22. The one or more non-transitory computer readable media of claim 19,

wherein the using the fault signatures stored in the fault dictionary and the failure log to identify a faulty scan cell in the scan chain comprises determining relative fault signature differences between the differential fault signature stored for the second scan cell and the complete fault signature stored for the first scan cell.

23. The one or more non-transitory computer readable media of claim 19, wherein the method further comprises storing an identity of the faulty scan cell.

24. One or more non-transitory computer readable media storing a series of instructions that are executable by a computer to perform a method of identifying a last fault in a faulty scan chain having multiple faults, the method comprising:

receiving a failure log indicating failures detected during scan chain testing of an integrated circuit, the failure log indicating failures on the faulty scan chain and failures on scan chains that are free of faults, the fault dictionary including one or more complete fault signatures and one or more incomplete differential fault signatures;

reading a fault signature from a fault dictionary that specifies failures that will occur in scan chains of the integrated circuit when a candidate scan cell in the faulty scan chain has a fault, the failures specified including failures in the faulty scan chain and failures on scan chains that are free of faults; and

comparing the failures specified in the fault signature for scan chains that are free of faults with failures in the failure log for scan chains that are free of faults to identify the candidate scan cell as the last fault in the faulty scan chain.

25. The one or more non-transitory computer readable media of claim 24,

wherein the method further comprises storing an identity of the candidate scan cell identified as the last fault.

26. The one or more non-transitory computer readable media of claim 24, wherein the method further comprises deriving the fault signature from at least one of the complete fault signatures and one of the differential fault signatures.

27. The one or more non-transitory computer readable media of claim 24, wherein the fault is a stuck-at 0 or stuck-at 1 fault.

28. The one or more non-transitory computer readable media of claim 24, wherein the fault is a timing fault.

29. The one or more non-transitory computer readable media of claim 24, wherein the failure log further indicates failures on primary outputs, wherein the fault signature further specifies failures that will occur at primary outputs when a candidate scan cell in the faulty scan chain has a fault, and wherein the act of comparing includes comparing the failures specified in the fault signature for the primary outputs with failures in the failure log for primary outputs to identify the candidate scan cell as the last fault of the multiple faulty scan cells in the faulty scan chain.

30. The one or more non-transitory computer readable media of claim 24, wherein the candidate scan cell is a last fault candidate scan cell, the method further comprising:

reading a fault signature for a first fault candidate scan cell;

creating a composite fault signature for the first fault candidate scan cell, the composite fault signature including failures defined in the fault signature for the first fault candidate scan cell that are upstream of the first fault candidate scan cell and failures defined in the fault signature for the last fault scan cell candidate that are downstream of the first fault candidate scan cell; and

comparing the failures in the composite fault signature with failures listed in the failure log to identify the first fault candidate scan cell as the first fault in the faulty scan chain.

31. The one or more non-transitory computer readable media of claim 24, wherein the method further comprises determining a first fault scan cell by comparing a composite fault signature for a candidate scan cell with failures listed in the failure log to identify the first fault scan cell in the faulty scan chain, wherein the composite fault signature for the candidate scan cell includes the fault signature for the candidate scan cell and its upstream scan cells specifying failures that will be detected in a defect-free scan chain as determined from a previously determined last fault scan cell and a fault signature of scan cells that are downstream of the candidate scan cell that will be detected in the scan chain that includes the first fault candidate scan cell.

32. One or more non-transitory computer readable media storing a series of instructions that are executable by a computer to perform a method of identifying a first fault in a faulty scan chain, the method comprising:

reading a fault signature for a candidate scan cell that may be the first fault in the faulty scan chain;

reading a fault signature for a previously determined last fault in the faulty scan chain, the fault signature being a differential fault signature from which a complete fault signature can be determined with reference to a base fault signature;

creating a composite fault signature for the candidate scan cell that includes failures defined in the fault signature for the candidate scan cell that are upstream of the candidate scan cell and failures defined in the fault signature for the last fault that are downstream of the candidate scan cell; and

comparing the failures in the composite fault signature with failures listed in a failure log to identify the candidate scan cell as the first fault in the faulty scan chain.

33. The one or more non-transitory computer readable media of claim 32, wherein the method further comprises storing an identity of the candidate scan cell identified as the first fault.

34. The one or more non-transitory computer readable media of claim 32, wherein at least one of the fault signatures for the candidate scan cell or the fault signature for the previously determined last fault is a relative signature, the method further comprising deriving the at least one of the fault signatures from at least a differential signature and a complete base signature.

35. The one or more non-transitory computer readable media of claim 32, wherein the fault is a stuck-at 0 or stuck-at 1 fault.

36. The one or more non-transitory computer readable media of claim 32, wherein the fault is a timing fault.