In-Circuit Test Effectiveness
In the previous section, the in-circuit test was deemed the most important for achieving the six sigma level. As shown in Table 4.8, significant savings in cost could be achieved if this test method could deliver PCBs directly to the customer, without having to undergo functional testing. A brief review of some of the terms and strategies of in-circuit testing are given to help in outlining a six sigma quality plan for testing. The plan is based on investigating the defect removal functions and rating their efficiency. This plan can also be used for any type of testing after assembly in manufacturing.
The functions that can be performed by in-circuit testing can comprise some or all of the following:
• Shorts and opens
• Polarity check
• Analog and digital component testing
• Analog, digital, and mixed signal in-circuit testing
• Analog, digital, and functional (powered-on) testing
• Digital pattern rate
• Interconnect and in-circuit boundary scan
The measures of a tester’s ability to correctly distinguish between bad and good PCBs are the test operation parameters: test coverage, bad test effectiveness, and good test effectiveness. They are measured as percentage values:
1. Test coverage (%): the test coverage for a given fault. Coverage of 〇 for a defect category means that this defect is not tested.
2. Bad test effectiveness (%): the percentage of bad components that fail a test. Thus, a tester with 100% bad test effectiveness will fail all bad items, whereas one with 0% bad test effectiveness will pass all bad items.
3. Good test effectiveness (%): the percentage of good parts that pass a test. Thus,a tester with 100% good test effectiveness will pass all good items, whereas one with 0% good test effectiveness will fail all good items.
4.4.4 Factors affecting test operation parameters
Factors that affect test effectiveness can be divided into three broad categories: technology, management decisions, and design for test (DFT) efforts. They are listed in Table 4.9 and further explained in the next section. A factor-based model could be created in order to make PGB design decisions during the development stage. The model could help the design team investigate the effect that different design choices would have on the test effectiveness.