Statistical Projection
Mathematical estimation of production losses during the manufacturing cycle allows for better planning and cost control in high volume electronics assembly. The yield drop modeling accounts for the cumulative effect of defects at each stage of the fabrication and assembly process. It uses historical data and process capability indices to predict how many boards will fail to meet the final acceptance criteria.
This approach moves beyond simple averages to identify the specific steps where the most major losses occur.
Variable Integration
Calculation of the expected output requires the input of defect rates from solder paste inspection and automated optical inspection. In yield drop modeling, the interaction between different process steps is analyzed to see if a small variation in one area leads to a large failure rate elsewhere. For example, a slight misalignment in screen printing might not cause an immediate failure but could lead to a higher rate of solder bridging after reflow.
The model incorporates these dependencies to provide a realistic view of the final pass rate.
Financial Impact
Knowing the probable number of scrap units helps a company set appropriate order quantities and pricing structures. Effective yield drop modeling ensures that a project remains profitable even when dealing with complex designs or new technologies.