| 摘要 |
The present invention is a method and instrument for testing a device. The device under test (DUT) may be an electronic device, circuit, PCB, or product. The present invention compares events measured on a known good DUT with events measured on a potentially faulty DUT. Events on the DUT may be stimulated by injecting one or more input signals into the DUT. Events are observed and measured at signal nodes termed "observation nodes." Events at the observation nodes are recorded and compiled into event lists. Event lists for a potentially faulty DUT are time aligned and compared with event lists for a known good DUT to determine whether the potentially faulty DUT is or is not actually faulty. The present invention can intelligently adapt the selection of observation nodes, on the basis of information about a DUT, to produce the most useful event lists for comparison. The present invention can backtrace through the event lists to determine the earliest or most upstream observation nodes at which the events of the potentially faulty DUT substantially differ from the events of the known good DUT. In this way, the cause of a fault can be isolated and repaired quickly. The present invention also can learn the fault cause and repair information in order to suggest the cause and needed repair when a similar event is observed on another DUT of the same type.
|
