Choose the feature and the question
Before requesting a section, state whether you are investigating landing geometry, copper distribution, an apparent interface defect, a dimensional concern, or a post-stress failure. Mark the exact structure and orientation. A cut through an adjacent via type may answer a different question even if the photograph looks similar.
Plan representative locations during initial release. If a destructive examination is likely, reserve appropriate samples or coupons rather than discovering after assembly that the only useful evidence is inaccessible.
Treat preparation as part of the evidence
IPC-TM-650 Method 2.1.1 describes microsection preparation. The preparation method and section plane affect what is visible, so the report should identify the method used and the sample's condition. Preparation artifacts must not be mistaken automatically for production defects.
Ask for an annotated image with scale, layer identification, feature identification, and the measurements needed for the stated question. A magnification label without an identifiable location is inadequate for comparing revisions or discussing a disputed result.
Use before-and-after comparisons carefully
For an illustrative investigation, compare an unstressed representative structure with a separate structure exposed to the agreed assembly simulation. Maintain sample identity and the exposure record. Differences can guide the investigation, but samples at different locations or with different geometry introduce alternative explanations.
- Document whether the sample came from a board or coupon.
- Identify the section direction and relevant layer pair.
- Retain original images as well as report annotations.
- Separate observations from proposed causes and acceptance decisions.
Combine geometry with performance evidence
A section shows a local structure at the time it is examined. Electrical monitoring can reveal behavior under selected conditions. When a reliability concern is intermittent or thermally dependent, coordinate the methods so the structural work investigates an electrically relevant sample instead of an arbitrary nearby feature.
Adjacent-layer access, one connection at a time
Dielectric · geometry + material
Core or build-up dielectric
Illustrative only. Copper thickness, dielectric construction and via spans require a reviewed stackup.
Common questions
Does one good section prove the entire lot is reliable?
No. It is evidence about a sampled structure. Lot acceptance and reliability conclusions require an agreed sampling and evaluation plan.
Is a test method also an acceptance specification?
Not necessarily. Define the applicable preparation method and the separate criteria used to judge the resulting measurements and observations.
References & further reading
Method & assumptions · General design guidance. Validate the proposed construction and acceptance criteria for your project.