VI Curve Tracing for Statistical Process Control in Electronics…

Beyond Pass/Fail: Using VI Data for Process Monitoring

Most VI curve testing is deployed as a binary pass/fail gate: the board either matches the golden reference within tolerance, or it does not. This is valuable, but it leaves significant diagnostic intelligence unused. The VI curve data collected during production testing contains process trend information that, when analyzed statistically, can detect manufacturing drift before defects occur.

This article describes how to use data from the FADOS MUX, the 100-Channel automated VI curve scanning system from ProT Ar-Ge, for statistical process control (SPC) in electronics manufacturing.

What VI Curves Reveal About Process Health

Every VI curve measurement captures the aggregate impedance of a circuit node. Changes in that impedance reflect changes in the components and connections that form that node. When the same test point is measured across hundreds or thousands of boards, the curve data forms a statistical population that reveals:

Solder quality trends: Increasing impedance scatter at BGA or fine-pitch QFP test points may indicate solder paste deposition drift, reflow profile changes, or paste aging
Component variation: Shifts in capacitor or resistor VI signatures across a production lot may indicate a component supplier change or incoming material variation
Assembly drift: Gradual impedance changes at specific test points can signal placement accuracy degradation, nozzle wear, or stencil clogging

Implementing SPC with FADOS Data

Step 1: Establish the Baseline

Run the FADOS MUX across a statistically meaningful sample of known-good boards, typically 30–50 boards from a stable production period. For each of the 96 test channels, the FADOS software records the VI curve. Extract a scalar metric from each curve for SPC tracking:

Impedance magnitude at a fixed frequency: the simplest metric, directly derived from the VI curve slope
Phase angle: distinguishes resistive from reactive deviations
Curve area: the enclosed area of the VI ellipse, sensitive to capacitance and ESR changes

Step 2: Calculate Control Limits

From the baseline dataset, calculate the mean and standard deviation for each test channel’s chosen metric. Set control limits using standard SPC practice:

Upper Control Limit (UCL): Mean + 3 sigma
Lower Control Limit (LCL): Mean - 3 sigma
Warning limits: Mean +/- 2 sigma

These control limits are specific to each test channel, a power rail test point has a different expected impedance range than a high-speed signal trace.

Step 3: Chart Production Data

As each production board passes through the FADOS MUX test station, plot the extracted metric for each channel on its respective control chart. Standard SPC rules apply:

Single point beyond control limits: Investigate immediately
Seven consecutive points on one side of the mean: Process drift, investigate
Trend of six consecutive increasing or decreasing points: Systematic drift, investigate
Two of three consecutive points beyond warning limits: Emerging shift, monitor closely

Step 4: Root Cause Investigation

When a control chart signals an out-of-control condition, the test channel identity provides direct diagnostic guidance:

Power rail channels drifting: Investigate solder paste volume at decoupling capacitor locations, check capacitor incoming quality
BGA channels showing increased scatter: Check reflow profile, inspect stencil condition, verify paste shelf life
Connector channels trending: Check component placement accuracy for connector pins
Widespread drift across multiple channels: Systemic issue, PCB fabrication change, new solder paste lot, reflow oven calibration drift

Practical Considerations for Indian EMS Facilities

Data Management

The FADOS software exports test data in standard formats suitable for import into SPC software or spreadsheet-based control charts. For facilities with MES integration, the data pipeline is: FADOS MUX → export file → MES database → SPC dashboard.

For smaller operations without dedicated SPC software, a structured spreadsheet approach works, extract the impedance metric per channel per board, calculate running statistics, and maintain control charts.

Test Station Calibration

For SPC data to be meaningful, the measurement system itself must be stable. Periodically run a reference board (a stable, gold-standard board stored specifically for calibration verification) through the MUX and verify that its measurements remain within expected limits. Any drift in the reference board’s readings indicates test system issues, probe wear, connector degradation, or MUX relay aging.

Operator Discipline

SPC requires consistent test execution. The board must seat correctly in the fixture every time, inconsistent contact pressure creates measurement noise that masks real process signals. Spring-loaded pogo pin fixtures with vacuum or mechanical hold-down provide the repeatability that SPC demands.

Connecting to Supplier Quality

When VI curve SPC data identifies incoming component variation as the root cause of drift, this data becomes a powerful tool for supplier quality discussions. Quantified impedance drift tied to specific component lots provides objective evidence for supplier corrective action requests.

The Business Case

For an EMS facility processing 500 boards per day through a FADOS MUX station, SPC analysis of VI curve data can:

Detect solder process drift before defect rates increase, avoiding scrap and rework costs
Identify component lot issues at incoming inspection, preventing contamination of production batches
Reduce defect escape rates by catching subtle deviations that pass simple pass/fail testing but indicate degraded assembly quality
Provide auditable process data for ISO 9001, IATF 16949, and AS9100 quality system requirements

Why Buy FADOS MUX from GSAS

GSAS Micro Systems is an authorized ProT Ar-Ge partner and ODM assembler. We support FADOS MUX deployments across India, from initial fixture design through SPC implementation, at our offices in Bengaluru, Hyderabad, Chennai, Pune, Mumbai, and Delhi NCR.