[previous] [next] [contents]    Control of internal resources and test program flow

The BS test controller model has to include a small number of internal resources and test program flow control mechanisms, which leads to the last set of basic test operations:

  • Load an internal counter with the required number of TCK cycles.
Operations like BIST or shifting a bit stream into the scan chain require the application of a pre-specified number of TCK cycles. Among other possible alternatives for this same purpose, the BS test controller model presented here assumes the existence of an internal counter that will be set to the required number of TCK cycles.
  • Select the error flag to be set if an expected value is not found.
Fault detection alone would not require any error flags, but their existence enables a limited fault diagnosis capability. If a new error flag is selected before the set of vectors for each step of the test protocol starts to be applied, it becomes possible to identify what type of faults were detected (even if not with what test vectors or in which cells). Assuming that this limited fault diagnosis capability is of interest, a basic test operation to select the required error flag becomes necessary.
  • Perform a conditional jump in the test program flow according to the value of the selected error flag.
Some types of faults might be potentially destructive, such as short-circuit faults, in which case it is recommendable to stop the test at once. This possibility requires the existence of a basic test operation controlling the test program flow, to enable corrective action with minimum delay.
  • Conclude the test program.
The simplest of all basic test operations, yet still indispensable.