Appeal No. 2004-2368
Application No. 09/395,854 Page 6
tester if latch 1 or latch 0 is bad, since they are both
downstream of the probe. As a result, this example (method) does
not teach “‘a re-detection of the defect by the
retesting...indicates that the first stage is non-defective, and
a detection of no defect...indicates that the first stage is
defective.’”
The last example provided by appellants (brief, page 8) uses
the same example of the two latches labeled 1 and 0. However, in
this example, the probe is moved from the input of latch 1 to the
output of latch 1. Once again, latch 1 is bypassed, but the
initial testing would have to have shown good data from latch 1
in order to move the probe to the right. Thus, this example does
not meet the claimed feature of first detecting a defect before
retesting the integrated circuit. Since latch 1's data output
was good, the result is that the claimed feature of "a detection
of no defect .... indicates that the first stage is defective" is
not met.
The examiner's position (answer, page 4 and 5) directs our
attention to the scan-chain probing disclosed in figure 5 of
Lindberg. In addition, the examiner creates an example (answer,
page 5) using Lindberg's method. In the example, the examiner
divides the sixteen latches into first and second circuit stages
Page: Previous 1 2 3 4 5 6 7 8 9 10 Next
Last modified: November 3, 2007