Page 128 - Building A Succesful Board-Test Strategy

114 BUILDING A SUCCESSFUL BOARD-TEST STRATEGY

generate fault simulations or conventional test programs, saving a significant
amount of work and time.
Many fault classes lend themselves to this type of detection, it can find solder
voids, misaligned or missing components, insufficient solder, shorts, and broken
connections with few false failures. Infrared inspection cannot identify joints with
excess solder, which most manufacturers regard as process faults and which X-ray
techniques will find, because the circuit appears to function normally.
Infrared also detects failures that x-ray misses. A cold solder joint, a faulty
ASIC, or an incorrect resistor, for example, could significantly change the board's
thermal signature, but would look no different in an x-ray image. In that respect,
the infrared technique resembles a functional test more than it does other forms
of inspection. The technique's supporters suggest placing it in the process after
in-circuit or other bed-of-nails test, possibly in place of functional test. This con-
figuration avoids adding a step that would lengthen the manufacturing process,
introduce another cycle of board handling, and possibly increase solder-joint
breakage and other handling-related failures.

3.3.4.2 Predicting Future Failures
Perhaps the most interesting aspect of the infrared approach is its ability to
detect latent defects—defects that do not affect the board's current performance,
but which may represent reliability problems after the product reaches customers.
Since the components still function and the board as a whole still works, these
faults generally defy conventional detection. Cracked solder joints, for example,
force the board current through a smaller-than-normal connection, creating a hot
spot that this tester will see. Some marginal components also fall into this fault
category.
To find such faults, manufacturers traditionally subject their boards to
some form of environmental stress screening (ESS), including burn-in, tempera-
ture cycling, and vibration, before performing an in-circuit or functional test.
The idea is to aggravate the latent faults until they become real faults and
therefore visible to subsequent test. Aside from requiring higher costs for equip-
ment, factory floor space, extra people, longer production times, and larger inven-
tories, ESS stresses good and bad components alike. Some authorities suspect
that such screening reduces overall board reliability and shortens board life.
In addition, vibration—the second most effective screen after temperature
cycling—is difficult to control and cannot apply stresses evenly across the
board, so results can be inconsistent. (ESS will be discussed in more detail in
Chapter 7.)
In contrast, stimulation during an infrared test subjects boards to no more
stress than they would experience in normal use. In one field test conducted by an
automotive manufacturer, the infrared test found all known failures from a sample
group of defective boards, whereas various versions of ESS revealed no more than
42 percent. In addition, the infrared system discovered that 2 percent of the boards
contained failures of which the manufacturer was unaware.