Testing and Diagnosis - Digital Microfluidic Biochips - page 105

2500

5000

Euler-path-based testing
Parallel scan-like testing

Euler-path-based testing
Parallel scan-like testing

2000

4000

1500

3000

1000

2000

500

1000

0 0

10 20 30 40 50 10 20 30 40 50

Array Size Array Size

Figure 4.16
Comparison of (a) testing complexity and (b) diagnosis complexity (additional steps) of a
parallel scan-like test and a Euler-circuit-based method.

Actual defect site

Candidate defect

Figure 4.17
Illustration of a candidate defect (incorrectly classified).

simply 1 − p . An electrode is a candidate-defective electrode if there is an
actual defect in either the same column or the same row as this electrode
(see Figure 4.17) . Therefore, the probability P cand that a defect-free electrode is
classified as a candidate defect site in an N × N array is given by:

P cand = (1 − p )(1 − (1 − p ) N−1 ) 2

For p << 1, and large N , we get

P cand = (1 − p )(1 − (1 − p ) N−1 ) 2

≈ p 2 (1 − p ) ( N − 1) 2

= p 2 (1 − p ) N 2

When diagonal testing is carried out, a defect-free electrode is classified as
a candidate electrode only if the following conditions hold: (1) there is a