1. Why SPI Alone Does Not Guarantee Quality

SPI systems measure solder paste:

• Volume

• Height

• Area

• Offset

• Bridging

Leading 3D SPI manufacturers such as Koh Young and CyberOptics have made volume measurement extremely accurate.

However, SPI only verifies printing quality, not final solder joint reliability.

Example:

• SPI shows OK paste volume

• Component placement slightly offset

• Reflow causes tombstoning

• AOI detects failure

Conclusion:
SPI = Process indicator
AOI = Result indicator

SPI alone cannot guarantee final product quality.

2. Why AOI Cannot Independently Determine Root Cause

AOI systems (such as those from Omron or Viscom) detect:

• Missing components

• Tombstones

• Polarity errors

• Solder bridges

• Insufficient solder

But AOI identifies symptoms, not root causes.

For example:

AOI detects insufficient solder.

Possible causes:

1. Low solder paste volume (printing issue)

2. Excess placement pressure

3. Poor reflow profile

4. Pad contamination

Without SPI data correlation, AOI alone cannot assign responsibility.

This is where closed-loop logic becomes essential.

3. SPI + AOI Responsibility Matrix Model

A responsibility matrix correlates SPI and AOI results:

This matrix eliminates subjective judgment and creates data-based root cause classification.

When implemented within MES or inspection software, the system automatically:

• Identifies trend patterns

• Detects repeated defect concentration areas

• Triggers process adjustment recommendations

This is the foundation of closed-loop optimization.

4. Closed-Loop Control Architecture

A typical closed-loop SPI/AOI architecture includes:

Step 1 – Data Acquisition

• 3D SPI captures paste metrics

• AOI captures post-reflow defect data

• PCB barcode links both datasets

Step 2 – Data Correlation

• Board-level mapping

• Component-level comparison

• Regional trend analysis

Step 3 – Process Feedback

Based on analysis, the system automatically adjusts:

To Printer:

• Squeegee pressure

• Print speed

• Separation speed

• Cleaning frequency

To Pick & Place Machine:

• X/Y offset correction

• Placement force optimization

Global equipment suppliers such as ASMPT and Hanwha Precision Machinery already support advanced data interfaces for this type of integration.

5. Real Case Study Example

Production Line: Automotive Control Board
Monthly Output: 120,000 boards

Before Closed-Loop Implementation:

• Average Solder Paste Cpk: 1.12

• AOI False Call Rate: 18%

• First Pass Yield (FTY): 91.5%

Major issue:
Localized low-volume printing at PCB corners causing tombstoning.

After Closed-Loop SPI→Printer Feedback:

• Automatic regional compensation enabled

• Cleaning frequency dynamically adjusted

• Placement offset fine-tuned based on SPI deviation

Results (3 months later):

• Solder Paste Cpk: 1.58

• AOI False Call Rate: 11%

• FTY: 96.3%

Defect escape rate reduced by 42%.

6. Cpk Improvement Comparison

The key improvement driver was not new hardware — it was data-driven feedback control.

Please wait for updates to the following related topics

✔ Converts inspection into process control
✔ Reduces engineering dependency
✔ Enables predictive maintenance
✔ Improves process stability
✔ Supports Industry 4.0 initiatives
✔ Enhances traceability and accountability