Page 112 - Building A Succesful Board-Test Strategy

98 BUILDING A SUCCESSFUL BOARD-TEST STRATEGY

Figure 3-1 Test and inspection represent a tradeoff. Extra effort expended at the
inspection stage should reduce the test burden downstream.

Creating inspection algorithms generally takes much less time than devel-
oping comparable test programs does. Also, most manufacturers place inspection
before test in the process. As a result, finding and repairing faults during inspec-
tion generally costs considerably less than uncovering and correcting those same
problems during test. In addition, inspection can find faults that defy electrical test.
insufficient solder, mousebites on traces, and other problems do not show up
during test because they do not directly affect board performance—yet. Sometimes
a solder joint may appear to make adequate contact even if it contains no solder
at all. These problems increase the likelihood that a previously functioning board
will fail in the field, after it has been shaken during shipping, for example. Repair-
ing boards before they leave the factory improves product reliability. Therefore, if
you can economically justify both inspection and test steps, shift as much as pos-
sible of the fault-finding burden to inspection.
At the same time, inspection cannot uncover many problems that show up
routinely during test. Inspection's ability to detect an incorrect part is extremely
limited. Test, however, can find exactly those problems. In-circuit and other bed-
of-nails techniques measure single components or clusters to determine if they are
correct and function properly. Functional, emulation, or system test examines the
behavior of the board (or system) as a whole.

3.1 Striking a Balance

Neither test nor inspection can find all faults for all manufacturing lines.
Certainly coverage overlaps, which can lead to the mistaken impression that one
technique or the other will suffice. But the areas of overlap are not sufficient. Each