DFT / ATPG · All levels

Boundary Scan Cells: Step-by-Step Walkthrough

Step-by-Step Walkthrough for Boundary Scan Cells.

Step-by-step analysis walkthrough

Use when you own Boundary Scan Cells in a DFT closure review.

  1. Open release criteria and failing metric dashboard.

  2. Partition failures by pattern family and operating context.

  3. Inspect constraints, clocks, masks, and unknown handling.

  4. Check architecture assumptions against implementation reality.

  5. Review diagnosis evidence for repeatability.

  6. Assign owner and propose smallest high-confidence change.

  7. Run timing, power, and quality regressions.

  8. Capture final signoff or waiver decision.

Artifacts to collect

  • boundary cell map, EXTEST report, pin coverage matrix

  • pattern tag manifest

  • tester program revision

  • signoff checklist

  • owner tracker

Decision memo template

diagram
DFT DECISION MEMO - Boundary Scan Cells
scenario:
metric:
hypothesis:
fix:
regression:
owners: DFT owner, package owner, board validation owner

Reference visuals

Boundary cell at IO

diagram
core logic <-> boundary cell <-> package pin

EXTEST drives/observes board interconnect independent of core paths.

DFT deep dive

Boundary scan and JTAG are board-level contracts, not just RTL features.

Concept diagram

diagram
JTAG ACCESS

TAP controller -> instruction register -> boundary/data register -> board test/debug

Metric graph

diagram
BOARD TEST READINESS

instruction coverage vs pin controllability

Reports and artifacts

  • TAP compliance report

  • boundary cell coverage matrix

  • EXTEST/INTEST results

  • debug lock policy log

Mini case study

Board bring-up blocked by pinmux override in one mode; TAP instruction decode and package table alignment fixed path.

Debug branches

  • Verify TAP state transitions

  • Audit package pin ownership

  • Check security lifecycle lock behavior

Senior review question

Ask: what evidence proves this DFT decision is safe for production?

Key takeaways

  • State metric, lot/corner context, and pattern tag with every claim.

  • Treat timing, power, and quality as one signoff problem.

Common pitfalls

  • Chasing coverage without legality checks.

  • Ignoring test-power side effects of pattern changes.

  • Debugging silicon without reproducible tags.

Principal DFT review addendum

Boundary scan cells wrap IO pins so interconnects can be tested independently from core logic using standard JTAG instructions.

Metric: boundary cell insertion coverage, pin-level controllability, EXTEST pass rate