Skip to main content

4.4 Programming & Tuning

Inspection programming translates raw component and solder data into actionable pass/fail judgments. The primary operational goal of tuning is to minimize the False Alarm Rate while ensuring zero escapes (False Negatives). This requires a structured approach based on Golden Samples, disciplined Library Management, and continuous iteration driven by Pareto Analysis of false calls. Inspection programs are not a substitute for process control; they function as a highly sensitive process monitor that flags upstream drift in printing or placement.

4.4.1 The Inspection System Programming Loop

Programming and tuning transform static design data into a robust, repeatable inspection judgment. This judgment must be consistent across shifts and operators.

  1. Data Import: CAD/Gerber data (centroid location, polarity, component shape) is imported and mapped to the machine's library.
  2. Teaching and Calibration: The machine is physically taught using the Golden Board (Chapter 2.5) to establish optimal lighting profiles (AOI) and slice planes (AXI/Laminography).
  3. Tuning: Thresholds and inspection criteria are iteratively adjusted to balance the detection capability against the operational cost of manual verification.

Mandate: The inspection program's success is measured not only by the defects it catches but also by the low number of false calls it generates.

4.4.2 Library and Data Standardization

A standardized library is essential for rapid changeovers and high throughput across multiple products.

  • One Package, One Library: A single library item should be created for each unique component package (e.g., 0402, 0.5 mm QFN). This item defines the Presence Window, Polarity Region, and Solder Pad ROI (Region of Interest) for inspection.
  • Reusability: Library items must be reused across all product programs to ensure consistency and minimize programming time for new products (NPI).
  • AXI Specifics: For AXI, the library item must lock parameters such as ball pitch, component height, slice plane locations (Laminography), and voiding rules (per ball vs. total area).
  • Golden Set: The program must be built using data derived from the Golden Board (known-good after FA) and, ideally, supplemented with Near-Limit Boards (assemblies with barely acceptable features) to teach the system what still passes.

4.4.3 Risk Classification and Limit Setting

Not all defects are equal. Inspection limits must be tailored to the risk level to prevent cosmetic issues from blocking critical high-reliability production (WIP).

Risk Class

Defect Type

Limit Strategy

Action upon Failure

Class A (Critical)

Missing Component, Polarity Reversal, BGA Bridge.

Tight Limits. Immediate hard stop or rejection to the rework station.

Must be addressed before the next panel begins.

Class B (Major Quality)

Tombstoning, Insufficient Fillet, Minor Skew.

Moderately Wide Limits. Requires verification and rework; data is logged for SPC (Chapter 4.5).

Used to monitor process drift.

Class C (Informational)

Smudges, Silk Screen Nicks, Flux Residue.

Chart Only. Never blocks WIP. Used for periodic housekeeping checks.

Data is captured but does not trigger an alarm.

4.4.4 Tuning and the False Call Reduction Loop

Tuning is the continuous process of adjusting parameters to reduce False Calls (False Positives) without introducing Escapes (False Negatives). This must be driven by data.

  1. Pareto Analysis: Run a sample lot (50–200 boards) and generate a Pareto chart of all false calls, classified by component reference designator (RefDes) and specific failure reason (e.g., "U10 fillet too reflective").
  2. Iterative Fixes: Address the top 3 false-call sources with the smallest possible change (e.g., adjust a single lighting angle, slightly widen one component's acceptable X/Y window). Avoid global threshold relaxation.
  3. Verification: Re-run a small mini-lot (10 boards) to confirm the false call rate drops and that the detection of true defects remains intact.
  4. Process vs. Program: If tuning cannot resolve a consistent false call (e.g., recurring Insufficient Solder calls on a specific component), the issue is likely Process Drift (low SPI volume, poor paste release) and must be referred to Manufacturing Engineering for a stencil or printer recipe fix (Chapter 1.5).

4.4.5 Change Control and Program Maintenance

Inspection programs must be treated with the same change control rigor as PnP programs (Chapter 2.2).

  • Version Control: The AOI/AXI program revision must increment with any change to the libraries, limits, or algorithm settings. The program version is tied directly to the product's Golden Recipe (Chapter 2.5).
  • Documentation: Every program change must include a one-line explanation in a changelog (e.g., "v2.1: Reduced false bridge calls on C47-C49 by softening lighting angle per QE approval").
  • Machine Learning (ML): If the system uses ML/AI for defect classification, the training set must be diverse, and a holdout set of images must be reserved for regular, independent verification of the model's accuracy.

Final Checklist: Inspection Program Health


Action

Control Point

Responsibility

Library Management

Component library items are reused, standardized, and version-controlled.

AOI/AXI Programmer

False Call Control

False call rate is measured weekly and falls below the target threshold (e.g., ≤ 1% per board).

Quality Engineer (QE)

Process Feedback

False call Pareto flags process drift (e.g., recurring shorts) instead of a tuning flaw.

Process Engineer (PE)

Recipe Integrity

Program revision, lighting/slice settings, and limits sheet are locked within the product's Golden Recipe bundle.

Manufacturing Engineering

Back to top