1.6 SPI Metrics, Cₚ/Cₚₖ, and Closed-Loop Rules

Solder Paste Inspection (SPI) is much more than a passive reject gate installed at the end of the line just to catch bad boards. It is a powerful process control tool that reveals exactly how the thermodynamic and mechanical variables of the printer are behaving in real-time. Using an SPI machine exclusively to sort boards into Pass or Fail bins neglects the majority of its engineering value.

The goal is converting raw SPI data into immediate, tactical engineering actions. The focus remains on active line control—interpreting the live statistical dashboard to stabilize the printing process before actual defects manifest.

The Vital Metrics: Volume is King

While modern SPI machines can measure dozens of geometric parameters, only three metrics truly drive process engineering decisions. The noise must be ignored, focusing exclusively on these three.

Metric	Engineering Definition	The Physical Consequence	Target Window
Volume (%)	The total cubic amount of paste deposited relative to the theoretical volume of the stencil aperture.	The single most critical metric. Volume correlates directly to final joint reliability. Low Volume guarantees opens or mechanically weak joints. High Volume guarantees bridging shorts and solder balls.	90% – 115% (Standard) 80% – 120% (Hard Limit)
Offset (X/Y)	How far the center of the paste deposit is shifted from the geometric center of the copper pad.	Indicates machine alignment drift, board clamping failure, or physical FR4 stretch. If the offset exceeds 25% of the pad width, the risk of bridging or tombstoning spikes exponentially.	± 0 µm (Target) < 30 µm (Warning)
Height (µm)	The absolute peak height of the solder deposit.	Used primarily to detect “dog ears” (spikes from fast separation) or “scooping” (dishing out of paste from high pressure). Height uniformity is critical for fine-pitch component placement consistency.	Stencil Thickness ± 10%

Pro-Tip: Do not obsess over the Area % metric. A print can have a perfect Area score, measuring 100% coverage of the pad, but actually be entirely insufficient in Volume if the paste was smeared far too thin. The process must always be controlled by measuring Volume.

Interpreting Cₚₖ on the SMT Line

Statistical expertise is not required to use Cₚₖ effectively on the shop floor. It serves directly as the machine’s stability score.

Cₚ represents Potential Capability, answering whether the machine is mechanically precise enough. If Cₚ is low, meaning the data shows a wide, scattered spread, the baseline process is unstable. In this case, intervention is required to check the squeegee pressure, look for paste viscosity breakdown, or check for insufficient board support.

Cₚₖ represents Real Capability, answering whether the machine is both centered on the target and precise. If Cₚ is high, showing a tight grouping, but Cₚₖ is low, the process is highly stable but off-target. A simple stencil alignment calibration or an offset adjustment in the printer software is likely required.

When analyzing the Cₚₖ score, anything above 1.66 is considered bulletproof, requiring no adjustments. A score between 1.33 and 1.66 is healthy, provided trend lines are monitored continuously. However, if the score drops below 1.33, the process is marginal and the risk of defects is becoming unacceptably high; the line must be stopped to make adjustments. If the score falls below 1.0, the process is fundamentally broken, actively producing scrap, and requires an immediate line stop.

Closed-Loop Rules: Printer ↔ SPI Feedback

Modern SPI machines can communicate directly with the printer via standard protocols to auto-correct process drift. This feature must be enabled, but clear limits must be established in the software to prevent the machine from continuously over-adjusting based on a single bad reading.

The Auto-Cleaning Trigger (Transfer Efficiency)

Instead of triggering the under-stencil wipe based on a fixed counter like wiping every 3 boards, the clean must be triggered based on measured Transfer Efficiency. A reliable guideline is that if the average Volume drops by more than 5% consistently across the last 3 consecutive boards, a Wet-Vac-Dry wipe must be triggered. This intelligent trigger cleans the stencil at the exact physical moment the microscopic apertures start to clog, which maximizes line cycle time without ever risking print quality.

The Alignment Auto-Correction

Similarly, if the X or Y offset drifts by more than 15 µm consistently in the same direction for 3 consecutive boards, an Offset Correction must be sent to the printer. Crucially, the maximum allowable auto-correction must be capped at 50 µm. If the drift suddenly exceeds 50 µm, it indicates a severe failure outside the bounds of software tuning, such as a board clamping failure, missing support pins, or a stretched stencil foil. At this point, the SPI must assert an alarm and halt the line.

Root Cause Mapping: From SPI Data to Physical Fix

When the SPI alarms, logical troubleshooting must be used to find the correct physical knob to turn on the printer, rather than guessing.

When the volume is consistently low, typically falling below 80%, the apertures are actively clogging or not filling during the print stroke. Begin by checking the paste roll. If the rolling bead is not at least 15mm in diameter, the hydraulic pressure is compromised, and paste must be added. Next, the squeegee speed must be observed. If the squeegee is moving too fast, typically over 80mm/s, the paste slides over the apertures instead of filling them, so it must be slowed down. Finally, the separation speed must be checked; if the table drops too fast, the paste cannot release and remains stuck inside the aperture, so the Z-axis drop must be slowed.

When the volume is consistently high, typically exceeding 125%, paste is leaking under the stencil foil, indicating the gasket seal has failed. First, the board support must be verified. When the center of the PCB is bowing down under the pressure of the squeegee, tooling pins must be added. Then, the clamping must be verified to ensure the foil clamps are holding the stencil perfectly tight against the PCB. Finally, the under-stencil cleaning effectiveness must be inspected; if the bottom of the stencil is heavily smeared with dried paste, it will prevent a flat seal.

Occasionally, volume that is excessively high in the center of the PCB but remarkably low on the edges may be observed. This symptom points to severe squeegee blade deformation. The metal blade is actively bowing like a bowstring under the applied downward force. Squeegee pressure must be drastically reduced or the entire blade assembly replaced immediately.

Final Checkout: SPI metrics, Cp/Cpk, and closed-loop rules

Key Parameter	Engineering Setting / Strict Rule
Volume Lower Spec Limit (LSL)	75% - 80% (Depends on the smallest pitch present)
Volume Upper Spec Limit (USL)	125% - 130%
Offset Limit	< 25% of the absolute pad width
Bridge Detect Threshold	150 µm (Detects thin smears connecting adjacent pads)
Critical Stop Rule	3 Consecutive Failures on the exact same reference designator.
Closed Loop Cleaning	ON (Triggered exclusively by Volume degradation trend).
Closed Loop Alignment	ON (Hard-cap the maximum allowable machine correction to 50 µm).