Hardware Reference
In-Depth Information
double propagation through gates 'c' and 'd' will cause the faulty values to re-
converge on the XOR gate 'e', thus cancelling each other. In other words, a classical
ATPG would not generate the vector #3 to detect a stuck-at-0 fault on node n 1 ,
whereas this vector allows the detection of a bridge-to-ground defect on node n 1
provided that the value of the bridge resistance falls into the interval [R 1 C ,R 3 C ].
2.3.2.3
Generalization
From the analysis conducted on the small illustrative example of the preceding sec-
tions, general comments can be drawn. An essential concept when dealing with
non-zero resistance bridging defects is the concept of critical resistance. Indeed al-
though the bridge resistance value is an unpredictable parameter that is not known a
priori, the critical resistance is a deterministic value that can be easily computed for
a given defect and a given test vector. Analogue Detectability Intervals can then be
calculated for different test vectors defining the range of detectable bridge resistance
values associated to the defect.
An important point is that Analogue Detectability Intervals are pattern depen-
dent: a specific ADI can be associated to each input vector taking into account both
the defect excitation and the effect propagation. As an illustration, Table 2.4 gives
the exhaustive list of Analogue Detectability Intervals associated to each input vec-
tor for the small defective circuit of Fig. 2.8 . Note that an empty interval is associated
to the vectors that do not excite or/and propagate the defect.
Despite its unpredictable parameter, defect detection can be optimized taking
into account the concept of ADIs. From the ADIs given in Table 2.4 , it appears
that six vectors are potentially able to excite the defect and propagate its effect to
Tabl e 2. 4
Associated ADIs ( Renovell et al. 1999 )
#
I 1 I 2 I 3 I 4
Associated ADIs
0
0000
Ø
1
0001
Ø
0
Œ0; R 1 C
2
0010
0
ŒR 1 C ; R 3 C
3
0011
4
0100
Ø
5
0101
Ø
6
0110
Œ0; R 2 C
0
7
0111
ŒR 2 C ; R 4 C
0
8
1000
Ø
9
1001
Ø
10
1010
Œ0; R 2 C
0
ŒR 2 C ; R 4 C
11
1011
0
12
1100
Ø
13
1101
Ø
R 1 C R 3 C R 2 C R 4 C
14
1110
Ø
15
1111
Ø
 
 
Search WWH ::




Custom Search