On this page
A ceramic via can pass a continuity check while still contributing excessive resistance or becoming intermittent under temperature and mechanical movement. The investigation must separate the via body from its face contacts, surrounding traces and measurement fixture. Begin with a defined electrical path and repeatable probe placement, then use physical inspection to answer the specific question raised by the electrical data. A dark center in a photograph or a single resistance reading cannot by itself identify incomplete filling, a crack or a poor external contact.
Key design decisions
- Define whether the measurement includes one via, a chain of vias or the entire circuit path.
- Keep force and sense locations fixed when comparing samples or process conditions.
- Document the original fault before applying heat, mechanical force or excessive test current that could alter it.
Identify every series contribution in the measurement
A two-terminal resistance reading can include test leads, probe contacts, printed traces, capture pads and one or more vias. Draw these segments and mark the physical measurement points. Without this map, a change attributed to via fill may actually be a different probe position or a longer conductor path.
If the circuit allows it, provide a dedicated structure that isolates the via connection from long face traces. For a via chain, record the number and arrangement of connections. Dividing total resistance by the number of vias is valid only after the other contributions are understood and the chain does not contain a dominant single defect.
Place voltage-sensing contacts around the intended segment
A four-wire arrangement separates current-carrying connections from voltage-sensing connections. This reduces lead-related error when sense contacts are positioned at the intended measurement boundary. On a small ceramic feature, the available pad area may limit how closely that boundary can approach the via.
Use a drawing or photograph to record force and sense positions on both faces. Avoid pressing so hard that the probe mechanically closes a crack or damages the metallization. Repeating a low reading under heavy probe force is not proof of an intact connection under normal use.
Check current dependence without changing the defect
Choose a test current that produces a measurable voltage without significant self-heating or electrical alteration of a weak contact. Record current and compliance settings with the resistance result. An instrument that reaches its compliance limit may not be applying the assumed current, making a simple voltage-over-current interpretation misleading.
Compare a small controlled range of currents when nonlinearity is suspected. A stable metallic path should behave differently from an intermittent interface, but current dependence alone does not identify the physical cause. Allow the sample to return to a consistent temperature between measurements and avoid a high-current pulse that could permanently modify the original condition.
Separate temperature coefficient from intermittency
A smooth resistance change with specimen temperature can be consistent with the conductor material, while sudden jumps may indicate a changing contact or crack. Measure specimen temperature rather than relying only on the chamber setpoint. Preserve the same electrical contacts through the comparison where practical.
Repeated thermal excursions can also change the fault. Record the order of measurements and whether a discontinuity appears during heating, cooling or a dwell. Compare identical temperature points on successive cycles. A value measured after cooling is not equivalent to an in-temperature measurement if the defect closes as the materials contract.
Use the pattern of evidence to narrow the cause
No single observation should carry the entire diagnosis. Electrical behavior, location and physical inspection should agree before a process change is selected. The table gives a sequence of questions rather than automatic failure labels.
| Observed behavior | Competing explanations | Discriminating check |
|---|---|---|
| Reading changes after probe reseating | Contact variation or local damage | Repeat with documented contact position and force |
| One chain has unusually high resistance | One weak via or longer face path | Segment the chain and compare trace contributions |
| Resistance changes with modest bending | Crack or fixture-contact movement | Immobilize external contacts and observe the local feature |
| Sudden jumps during temperature change | Intermittent interface or lead motion | Correlate specimen temperature and fixed-contact measurement |
| Open circuit with visible central void | Incomplete path or unrelated surface appearance | Section the electrically implicated region |
Inspect the electrically implicated region
Begin with both faces and the capture-pad boundaries. Check whether registration, a cracked pad or an interrupted trace could explain the result before sectioning the via. A top view can show an opening or depression but cannot reveal the complete conducting path through the thickness.
When destructive analysis is justified, locate the section from the electrical evidence. A section that misses the weak region can make an intermittent via appear sound. Record preparation direction and examine whether polishing introduced pullout or cracks. Compare the section with pre-preparation images so an artifact is not mistaken for the original defect.
Use a simple resistance model only as a consistency check
A uniform cylindrical conductor can be approximated from resistivity, length and cross-sectional area. Real filled or wall-metallized vias may have nonuniform geometry and interfaces, so the model is a starting comparison rather than a released current-carrying rating. Specify which construction is being modeled.
If a measured path is much higher than the estimate, investigate contact contributions, actual conducting area and material state before changing the nominal hole diameter. Conversely, agreement with the model does not establish resistance stability under cycling. Initial resistance and long-term integrity require different evidence.
Rvia ≈ ρ × L / Aconducting
- ρ: resistivity of the actual processed conductor
- L: effective path length through the substrate
- Aconducting: effective conducting cross-section
Uniform ohmic conduction with negligible interface resistance; not applicable as a direct description of cracked, partially filled or unknown constructions.
Tie the next experiment to a specific mechanism
A useful corrective comparison changes one suspected contributor while preserving the others. Examples include improving probe access, comparing hole preparation, checking filling continuity or evaluating capture-pad registration. Retain an unchanged comparison group so day-to-day measurement drift does not appear to be a process improvement.
The final record should identify the fault location, measurement boundary, sample history and physical observations that support the conclusion. State what remains unexplained if the evidence does not distinguish two mechanisms. This prevents a cosmetic improvement in the via surface from being accepted as proof that the intermittent electrical path has been removed.
Provide the via measurement and construction data
Send the electrical path and physical context needed to localize the problem.
- Substrate thickness, hole geometry, stated fill or wall-metallization construction and both-face artwork.
- Measurement-point photographs, force/sense arrangement, test current, compliance and specimen temperature.
- Resistance histories, intermittent-event conditions and the identity of affected panel positions.
- Pre-test images, process sequence and any sections or inspection records tied to the measured fault.
The drawing-upload form loads as you reach this section.