A semiconductor package manufacturing change can affect more than where a component is assembled. Differences in production site, materials or process conditions may raise questions about package dimensions, solderability and long-term reliability. A documented assessment of a QFP assembly-site transition illustrates how these concerns can be checked before parts enter production.
The change covered QFP15 100-pin and 128-pin packages. Assembly moved from Singapore Epson Industrial (SEP) to Advanced Semiconductor Engineering, Inc. Chung-Li (ASECL). The wafer process, electrical and visual test process, and packing and shipping process were not changed. The assessment therefore focused on the assembly-related differences and whether the new parts remained suitable for existing production handling.
What changed in QFP package assembly
The lead-frame material remained copper, but the lead-frame dimensions were adopted from the subcontractor’s standard. The mold compound changed from the CV8000 series to the EME-G600 series. Both compounds were specified as halogen-free and rated UL-94 V-0 for flammability. IC chip thickness changed from 400 µm to 300 µm, using the subcontractor’s standard thickness and process conditions.
These differences define the scope of the engineering change. They do not, by themselves, establish whether the new assembly is equivalent in use. That requires checking relevant characteristics against the existing parts, including dimensional tolerances, lead strength, solderability and reliability.
How the semiconductor package manufacturing change was evaluated
The verification plan connected each change to a corresponding check. Lead-frame dimensions were assessed using measurement data; lead strength and solderability were tested; and reliability testing was applied to the mold compound and the thinner chip configuration. The evaluation reported passing results for the identified verification items, with no problems found.
Dimensional capability and appearance
For both package types, the comparison covered package length and width, overall dimensions, seated and package heights, terminal width and thickness, terminal length, coplanarity, and other terminal geometry. The reported measurements met the specified requirements. The assessment also stated that all reported dimensional Cpk values were above 1.67.
Photographs showed a difference in the package index mark between the current and new parts. The assessment identified no corresponding change to external dimensions or their tolerances. For engineering and receiving teams, this distinction matters: an appearance difference may require recognition in inspection documentation, but it should not be mistaken for a change in package fit when the dimensional results show otherwise.
Reliability, lead strength and solderability
Reliability tests were performed under the same conditions used for the current parts. The test set included high-temperature bias, high-temperature and high-humidity bias, high-temperature storage, temperature cycling, pressure cooker exposure and resistance to soldering heat. The listed tests recorded zero failures and were judged to pass.
Separate checks addressed mechanical and assembly-related properties. Lead pull and bending tests passed, as did solderability tests following two specified aging conditions. The comparison also reported the same resistance to soldering heat, described as the same moisture sensitivity level (MSL), for the current and new parts.
The change in chip thickness was included in the reliability evaluation. The assessment concluded that reducing thickness from 400 µm to 300 µm had no influence on the IC chip itself. This conclusion is specific to the evaluated parts and test results; it should not be generalized to other package designs or process changes without their own review.
What the results mean for production teams
For system integrators, manufacturing engineers and plant operations teams, the practical question is whether a supplier or assembly-site change requires adjustments to established production handling. In this case, the assessment concluded that the engineering change presented no concern and that the parts could be used with the current handling method, without adjustment.
That conclusion rests on the documented scope and evidence: unchanged wafer, test and shipping processes; verification of the assembly-related changes; dimensional results within the stated specifications; and passing reliability, lead-strength and solderability checks. The assessment also noted that the subcontractor had produced more than 1,000 million pieces under the described standard conditions since 2009. This production history provides context, while the product-specific verification remains central to the change decision.
When managing a similar supplier or site transition, teams can use this approach to organize review:
- Identify which manufacturing steps and product characteristics are changing, and which remain unchanged.
- Link each identified change to an appropriate inspection, measurement or test.
- Compare external dimensions and tolerances, not just package appearance.
- Review reliability and assembly-related results against the established parts or requirements.
- Document whether existing production handling remains applicable.
Conclusion
A controlled semiconductor package manufacturing change depends on clear change boundaries and evidence tied to the affected characteristics. In this QFP case, the assembly site, mold compound and chip thickness changed, while other production stages remained unchanged. Dimensional comparisons and reliability-related tests passed, supporting continued use of the parts with the existing handling method. For manufacturers and integrators, the broader lesson is to make change approval a traceable engineering decision: define what is different, verify the risks that follow from those differences, and record the basis for production acceptance.