Direct answer
Two things matter: whether the laboratory has both non-destructive and destructive analysis methods (X-ray, CT, C-SAM, dye penetrant, cross-section, SEM/EDS), and whether it can deliver a failure mechanism judgement rather than photographs alone. A laboratory with only X-ray and no cross-section or composition analysis usually never reaches the root cause.
Four levels of analysis — miss one and the chain breaks
- Appearance and electrical confirmation: first confirm the failure can be reproduced and record the deviation in electrical parameters. Skip this and everything afterwards is guesswork.
- Non-destructive inspection: X-ray for voids and cracks inside solder joints, CT for three-dimensional structure, C-SAM for delamination.
- Destructive analysis: cross-section for metallography and IMC (intermetallic compound) thickness, dye penetrant for the crack path, decapsulation for the die surface.
- Composition and structure analysis: SEM/EDS for elemental composition and micro-morphology, FTIR for organics, ion chromatography for residual ionic contamination.
With only the first two levels the conclusion usually stops at "there is an anomaly here"; only levels three and four can answer "why did it happen". The value gap is wide — the former lets you replace a part, the latter lets you change the design.
Six high-frequency failure modes in automotive PCBA
- Thermal fatigue cracking of solder joints: under temperature cycling the thermal expansion of solder and component mismatch, and cracks start at the joint edge and propagate through
- Tin whisker shorts: pure tin plating grows whiskers under compressive stress, especially dangerous on fine-pitch devices
- Electrochemical migration (ECM): moisture plus bias plus ionic contamination drives metal ions to migrate and form dendrites
- Connector terminal fretting corrosion: vibration causes small sliding at the contact surface, oxidation raises contact resistance
- Corrosion beneath conformal coating: local coating defects form corrosion cells — hidden and hard to locate
- CAF (conductive anodic filament): conductive paths form inside the PCB substrate along glass fibre bundles
Three things to state clearly when submitting samples
- The failure symptom: when it fails, under what conditions, and in what proportion
- The operating environment: temperature, humidity, vibration, supply conditions and time in service
- What has already been checked: avoids duplicate work and helps judge the direction
Also provide where possible: known-good units from the same batch (for comparison), the schematic or at least the critical nets, and process information (soldering method, cleaning process, coating).
What the report should look like
A usable failure analysis report contains at least: a record of reproducing the failure, the analysis methods and steps, key images (with scale and location marked), composition or structure data, a judgement of the failure mechanism, and improvement recommendations.
If the report is a pile of photographs with no mechanism conclusion and no corrective direction, it is an inspection record, not a failure analysis.
Where we stand
TMC LAB holds both CMA and CNAS accreditation. Failure analysis covers X-ray, CT, C-SAM, cross-section, SEM/EDS and dye penetrant, with case experience on solder joint, plating and migration failures in automotive PCBA. Send us the failure symptom and what has already been checked, and we will decide which analysis path applies first.
Related pages for this question
Last updated Technical criteria follow the current standards and the agreed scope; please confirm the specific project with the lab.
