HDI Reliability Engineering Guide
Review finished geometry, voltage and laminate integrity before defining CAF risk
Conductive anodic filament failure can gradually reduce insulation resistance between conductors inside an HDI PCB. The risk is not controlled by spacing alone. It develops through the combined influence of finished conductor geometry, long-term electrical bias, moisture exposure, laminate-interface integrity and fabrication quality.
Does the finished conductor geometry provide sufficient insulation margin for the real voltage, environment and service-life requirement?
Small finished spacing, sustained DC bias, humidity, resin-to-glass weakness and process-induced laminate damage.
A project-specific DFM, material-control and qualification plan agreed before prototype or production release.
1. Understand How Conductive Anodic Filament Failure Develops
Conductive anodic filament, or CAF, is an internal electrochemical migration mechanism in which a copper-containing conductive path develops through the PCB laminate between conductors at different electrical potentials.
The path frequently develops along a weakened resin-to-glass interface or another susceptible region inside the laminate. Under humid conditions, absorbed moisture can support ionic conduction. When a sustained DC electrical field is present, copper ions can migrate from the positively biased conductor toward the negatively biased conductor.
As the conductive path develops, insulation resistance falls. The electrical result may initially appear as increased leakage or intermittent malfunction. Continued filament growth may eventually produce a permanent short circuit.
Moisture uptake
Moisture enters the laminate or accumulates along a susceptible internal resin-to-glass interface.
Internal path formation
Material weakness, drilling damage, local voiding or interface degradation creates a possible migration path.
Copper-ion migration
A sustained electrical field drives electrochemical migration through the moisture-assisted internal path.
Insulation degradation
The conductive filament develops until leakage increases or an internal short occurs.
Key engineering principle
CAF generally requires several conditions to act together: moisture, an electrical potential difference, a source of mobile copper ions and a susceptible internal migration path. Removing or controlling one factor can reduce risk, but qualification must consider the complete system.
2. Separate CAF from Surface Migration and Microvia Cracking
CAF is sometimes grouped with other HDI reliability failures, but its location, physical mechanism and electrical result are different.
| Failure Mode | Typical Location | Primary Driving Factors | Typical Electrical Result |
|---|---|---|---|
| Conductive anodic filament | Inside the laminate dielectric | Moisture, electrical bias and a susceptible internal path | Leakage or short circuit |
| Surface dendritic growth | PCB surface | Ionic contamination, moisture and electrical bias | Surface leakage or short circuit |
| Microvia corner cracking | Microvia-to-capture-pad interface | Thermal-mechanical stress and copper fatigue | Intermittent or permanent open circuit |
| Barrel separation | Plated-hole copper structure | CTE mismatch, repeated thermal stress and interface weakness | Open circuit |
| Delamination | Resin, copper or glass interface | Moisture, thermal stress or insufficient bonding | Structural damage and possible electrical failure |
Microvia cracking is primarily a thermal-mechanical copper-interconnect failure. CAF is an electrochemical insulation failure. Both can occur in a complex HDI product, but they require different test methods, failure-analysis evidence and corrective actions.
For thermal-mechanical failure identification, review microvia corner cracking versus barrel separation in HDI PCBs .
3. Why HDI Structures Require More Careful CAF Risk Control
HDI technology does not inherently create CAF. The challenge is that higher interconnect density reduces the available margin for geometric, material and fabrication variation.
Reduced conductor separation
HDI designs commonly include:
- Dense via and plated-through-hole arrays
- Reduced via-to-via spacing
- Reduced via-to-plane clearance
- Fine inner-layer conductor spacing
- Thin dielectric layers
- High-density BGA escape routing
- Multiple voltage domains within a compact area
As spacing decreases, drill-position variation, plating thickness, finished-hole diameter and layer registration become a larger percentage of the available insulation distance. A structure that appears acceptable in nominal CAD geometry may provide much less dielectric margin after manufacturing tolerances are applied.
Sequential lamination complexity
Advanced HDI boards may use multiple lamination cycles. Each cycle introduces variables that must remain controlled:
- Prepreg resin flow and local resin distribution
- Material moisture condition before lamination
- Local copper balance
- Registration stability
- Resin-to-glass adhesion
- Internal void formation
- Accumulated thermal history
Avoid an incorrect conclusion
Multiple lamination cycles do not automatically cause CAF. They increase the number of material and process variables that must be controlled, verified and documented.
Dense via fields
Microvias themselves are not the direct cause of CAF. The review must consider the complete structure around them, including finished conductor spacing, adjacent voltage domains, surrounding laminate, glass orientation and fabrication tolerances.
A dense via field can combine closely spaced plated conductors, reduced dielectric margins, several electrical potentials and increased sensitivity to drilling and registration variation.
High-reliability and harsh-environment applications
CAF becomes more significant when dense geometry is combined with long-term electrical bias, moisture, temperature variation or extended service life.
| Application Environment | Typical Exposure | Recommended Review Focus |
|---|---|---|
| Aerospace and defense | Long service life, environmental variation and high failure consequence | Material control, traceability and project-specific qualification |
| Automotive electronics | Humidity, temperature cycling and extended qualification life | Finished spacing, material performance and representative testing |
| Outdoor communications | Humidity, possible condensation and continuous electrical bias | Environmental stress and insulation stability |
| Industrial controls | Long operating time, contamination and temperature changes | Geometry, cleanliness and material qualification |
| High-voltage controls | Greater potential difference across the dielectric path | Voltage-domain and minimum-distance review |
| Long-life medical electronics | Extended field life and high cost of latent failure | Material stability, process evidence and qualification coverage |
4. Review Finished Conductor Spacing—not Only Nominal Hole Pitch
Spacing is an important CAF variable, but a nominal PCB design-rule value cannot predict long-term reliability by itself.
Relevant dimensions may include:
- PTH-to-PTH finished hole-wall spacing
- Via-to-via spacing
- Via-to-plane spacing
- Via-to-inner-layer conductor clearance
- Trace-to-trace spacing
- Z-axis dielectric separation
- Dense drill-field geometry
- Conductor orientation relative to the glass weave
Why nominal center-to-center pitch can be misleading
Designers often evaluate drilled-hole center-to-center pitch. CAF margin is more directly influenced by the finished dielectric distance between plated conductors.
The actual remaining laminate can be affected by:
- Finished hole diameter
- Copper plating thickness
- Drill-position tolerance
- Layer registration
- Local hole-field density
- Resin distribution around the drilled structure
Worst-case finished-spacing review
Compare the nominal geometry against finished copper dimensions, drill-position tolerance, layer-registration tolerance, plating variation and any other project-specific process allowance. The objective is to identify the smallest realistic copper-to-copper insulation distance.
Two designs with the same nominal hole pitch can therefore have different CAF margins. A larger finished hole, heavier plating or unfavorable drill offset can reduce the laminate remaining between adjacent plated conductors.
Why a fixed minimum spacing is misleading
A spacing value cannot be classified as universally safe without defining:
- The voltage difference between adjacent conductors
- The laminate and prepreg construction
- Humidity and temperature exposure
- The required product lifetime
- Finished manufacturing tolerances
- The conductor and glass-weave orientation
- The test duration and test structure
- The applicable acceptance criteria
Do not publish one universal CAF-safe spacing
A geometry that performs adequately in a controlled indoor product may not provide sufficient margin in an outdoor, automotive, aerospace or high-voltage application.
Recommended spacing review sequence
Identify the minimum finished distance
Review copper-to-copper separation after plating, drill-position and registration tolerances are applied.
Map the voltage domains
Identify adjacent conductors that remain at different DC potentials during normal and standby operation.
Apply the environment and lifetime
Review expected humidity, temperature, operating duration and reliability class.
Define representative qualification
Determine whether a standard CAF vehicle represents the product or whether a custom coupon is required.
5. Evaluate Voltage Bias Together with the Dielectric Path
Voltage alone does not determine CAF susceptibility. Risk depends on the interaction between the potential difference, the effective dielectric path, moisture exposure and internal laminate condition.
When the same voltage is applied across a smaller distance, electrical-field stress across the dielectric path increases. If moisture and a susceptible internal interface are also present, electrochemical migration can be accelerated.
Long-term DC bias matters
CAF-sensitive structures can include:
- Power-to-ground plated-hole patterns
- Always-on circuits
- Standby power networks
- Adjacent high- and low-voltage domains
- Permanently biased backplanes
- Outdoor communication systems operating continuously
A short functional test may confirm immediate electrical operation, but it does not reproduce the accumulated effect of years of bias under humid conditions.
Short-term insulation tests are not CAF qualification
| Test | Primary Purpose | What It Does Not Demonstrate |
|---|---|---|
| Hi-pot or dielectric withstand | Short-duration insulation withstand | Long-term internal electrochemical migration resistance |
| Insulation resistance | Resistance between conductors under defined conditions | Full product-life CAF resistance |
| SIR testing | Surface insulation and contamination-related behavior | Internal laminate CAF performance |
| CAF testing | Internal electrochemical migration susceptibility | Thermal-mechanical via or microvia life |
| IST or thermal cycling | Thermal-mechanical interconnect reliability | Internal electrochemical migration resistance |
| Microsection | Physical construction and defect examination | Long-term electrical behavior by itself |
6. Include Humidity, Temperature and Exposure Time in the Risk Model
Moisture is an enabling condition for CAF because it supports ionic transport within a susceptible internal laminate path.
What moisture changes
- Increases ionic mobility
- Reduces internal insulation stability
- Exposes weaknesses in the resin-to-glass interface
- Interacts with ionic residues or material impurities
- Accelerates electrochemical migration under electrical bias
The risk depends not only on the laminate’s published moisture-absorption value, but also on material storage, drying, lamination, hole-wall condition and the final operating environment.
What temperature changes
- Moisture diffusion rate
- Electrochemical reaction rate
- Resin-to-glass interface stability
- Insulation-resistance measurement behavior
- Accelerated-test severity
- Long-term laminate aging
Application conditions determine qualification
A temperature-controlled indoor device, an outdoor communication system and an aerospace control unit should not automatically use the same risk classification or test plan.
Conformal coating does not eliminate internal CAF risk
Conformal coating mainly protects the PCB surface from moisture and contamination. It may reduce environmental exposure, but it cannot repair:
- Internal laminate damage
- Resin-to-glass separation
- Insufficient internal conductor spacing
- Voids created during lamination
- Drilling damage around plated holes
CAF is an internal laminate failure mechanism. Surface coating is not a substitute for internal geometry control, material integrity and representative qualification.
7. Review CAF Data, Resin-Glass Integrity and Material Substitution Risk
Material selection is central to CAF risk control, but the term “CAF-resistant laminate” should not be interpreted as “CAF-proof.”
CAF-resistant does not mean risk-free
A laminate may be formulated or qualified to provide improved resistance to conductive filament growth. Final board performance still depends on:
- Actual finished PCB geometry
- Drilling and hole-wall quality
- Lamination conditions
- Material moisture handling
- Electrical bias
- Manufacturing cleanliness
- The final operating environment
Material qualification data should support project-level engineering review. It should not replace evaluation of the finished board construction.
Resin-to-glass interface integrity
CAF frequently follows a weakened resin-to-glass interface. Relevant material and fabrication variables include:
- Resin chemistry
- Glass-fiber treatment
- Resin-to-glass adhesion
- Moisture resistance
- Mechanical drilling stress
- Lamination quality
- Material aging
- Repeated thermal exposure
A higher Tg does not automatically mean better CAF resistance
Glass-transition temperature is important for thermal performance, but it is not a complete CAF qualification metric.
A laminate review may also need to consider:
- CAF qualification data and test conditions
- Moisture absorption
- Resin chemistry
- Glass treatment and glass style
- Z-axis expansion behavior
- Prepreg resin content
- Lamination-process compatibility
- Sequential-lamination behavior
Material substitution risk
Two laminates with similar Tg, Dk, Df, thickness and flame rating can still use different resin systems, glass treatments, prepreg constructions or CAF test conditions.
| Material Review Item | Why It Matters | What to Confirm |
|---|---|---|
| CAF performance data | Shows resistance under defined structures and conditions | Whether the data represents the project geometry and environment |
| Moisture absorption | Influences insulation stability and environmental sensitivity | Suitability for the intended operating environment |
| Resin and glass system | Influences internal interface integrity | Material construction and supplier qualification data |
| Prepreg construction | Influences resin flow, fill and local dielectric quality | Resin content, glass style and lamination compatibility |
| Sequential-lamination compatibility | Repeated cycles increase the accumulated process history | Approved construction and process window |
| Material substitution | Similar Tg or Dk does not guarantee equivalent CAF behavior | Customer approval and requalification requirements |
Avoid approval by datasheet comparison alone
A proposed alternate laminate should be reviewed for CAF performance, moisture behavior, resin and glass construction, available prepregs, lamination compatibility and customer qualification requirements.
8. Control Drilling, Desmear, Lamination and Finished Geometry
A laminate with strong published CAF performance can still be compromised by fabrication damage, moisture or uncontrolled finished geometry.
Drilling-induced laminate damage
Mechanical drilling can disturb the material surrounding a plated through-hole. Relevant variables include:
- Drill-bit condition and tool life
- Feed and speed
- Panel stack height
- Heat generation
- Glass-fiber fracture
- Resin smear
- Hole-wall roughness
- Dense-hole-field stability
The objective is not only to produce an electrically continuous plated hole. The surrounding dielectric structure must also remain sufficiently intact.
Desmear and hole-wall preparation
Insufficient desmear may leave resin residue that affects metallization. Excessive treatment may damage the resin system or expose glass fibers unnecessarily.
The process window should support:
- Reliable copper adhesion
- Controlled resin removal
- Stable hole-wall geometry
- Limited resin-to-glass interface damage
- Consistent metallization and plating
Lamination quality
Potential lamination-related risks include:
- Internal voids
- Insufficient resin flow
- Local resin starvation
- Moisture before lamination
- Unstable pressure or temperature profiles
- Incompatible material combinations
- Uneven copper distribution
- Uncontrolled repeated thermal history
Finished-hole geometry and plating
Copper plating changes the actual conductor geometry after drilling. The remaining laminate between adjacent plated holes can be affected by:
- Finished hole diameter
- Plating thickness
- Drill-position variation
- Layer registration
- Local plating distribution
Use the correct manufacturing question
Do not ask only, “What is the drill-to-drill spacing?” Ask, “What is the minimum finished conductor-to-conductor distance after plating, drill-position and layer-registration tolerances are applied?”
Cleanliness and ionic contamination
Ionic contamination is more commonly associated with surface insulation resistance and surface electrochemical migration. However, process cleanliness remains part of a high-reliability PCB control system.
Not every humidity-related insulation failure should be called CAF. Failure analysis should determine whether the conductive path is on the surface, inside the laminate, associated with a drilled-hole interface or related to assembly contamination.
| Process Stage | Potential Risk | Recommended Evidence |
|---|---|---|
| Material storage | Moisture absorption before lamination | Storage, handling and baking controls |
| Lamination | Voids, resin starvation or unstable interfaces | Press controls, material records and microsection evidence |
| Drilling | Hole-wall or resin-to-glass interface damage | Tool management and hole-wall inspection |
| Desmear | Under- or over-treatment | Chemical-process control and qualification records |
| Plating | Finished-geometry and insulation-margin variation | Copper thickness and finished-hole measurements |
| Final cleaning | Ionic residue and surface insulation risk | Cleanliness and related verification requirements |
| Qualification | Coupon does not represent the real product | Coupon geometry and test-plan review before release |
9. Qualify CAF Risk with a Representative Test Structure and Plan
IPC-TM-650 Method 2.6.25C is used to evaluate the propensity for conductive anodic filament growth and related internal electrochemical migration mechanisms. The test plan must still reflect the specific product and qualification objective.
A standard CAF vehicle can support laminate or process comparison. A critical HDI design may require a product-representative coupon that reflects the actual conductor geometry, finished spacing, voltage relationship and board construction.
Five questions the qualification plan must answer
Which geometry is being evaluated?
Define whether the risk path is hole-to-hole, hole-to-plane, conductor-to-conductor, Z-axis or another product-specific structure.
Does the spacing represent the product?
Confirm that the coupon includes a finished spacing and construction relevant to the actual HDI design.
Is the applied bias relevant?
Material screening, process qualification and application simulation may require different electrical conditions.
Are the environmental conditions appropriate?
Define temperature, humidity, exposure time and preconditioning based on the purpose of the qualification.
How will failure be defined?
Establish baseline resistance, monitoring frequency, resistance-drop criteria, treatment of intermittent events and post-test analysis.
Standard and product-representative coupon approaches
| Coupon Approach | Recommended Use | Main Limitation |
|---|---|---|
| Standard CAF test vehicle | Material or manufacturing-process benchmarking | May not represent the product’s minimum finished geometry |
| Product-representative coupon | Evaluation of actual geometry and electrical-risk conditions | Requires project-specific design and approval |
| Lot-qualification coupon | Production or process monitoring | Must remain aligned with the approved manufacturing process |
| Failure-analysis coupon | Reproducing or isolating a suspected field mechanism | May require several structures and stress conditions |
| Customer-specific coupon | Aerospace, automotive, defense or high-voltage programs | Conditions and acceptance criteria require agreement |
How test results should be interpreted
A complete evaluation may review:
- Initial insulation resistance
- Resistance trend over time
- Sudden or progressive resistance reduction
- Intermittent leakage events
- Correlation between test channels
- Post-test microsection evidence
- Additional material or chemical analysis when required
What CAF testing cannot prove
- It cannot independently prove every product configuration is failure-free.
- It cannot replace thermal cycling or microvia reliability testing.
- It cannot replace SIR or assembly-cleanliness evaluation.
- A standard coupon may not represent the real product geometry.
- A passing laminate datasheet does not qualify every finished PCB design.
- A failed result still requires structured root-cause analysis.
Do not extend one passing result beyond its scope
A standard coupon passing does not automatically qualify a product with different spacing, voltage, laminate construction, manufacturing process or environmental requirements.
For complementary HDI reliability methods, review HDI thermomechanical failure testing .
10. HDI PCB CAF Risk Review Checklist
The design package should allow the PCB manufacturer to connect the electrical environment to the actual stackup, materials, finished geometry and qualification requirements.
Design review
Material review
Fabrication and qualification review
11. Supplier Evaluation Questions for CAF-Sensitive HDI Projects
A PCB supplier should not answer every CAF question with a laminate datasheet. Reliable control requires coordination between design, material selection, fabrication, qualification and change control.
| Supplier Question | Why It Matters | Expected Evidence |
|---|---|---|
| Which laminate systems have relevant CAF data? | Confirms the material was evaluated under defined conditions | Supplier data, qualification reports or approved specifications |
| How is finished hole-wall spacing calculated? | Nominal drill pitch may overstate the actual insulation margin | Finished geometry and tolerance review |
| Are drill and registration tolerances included? | Worst-case conductor spacing may be smaller than nominal | Stackup-specific manufacturability review |
| How are dense drilled-hole fields controlled? | Drilling damage can create susceptible internal paths | Tool, process and inspection controls |
| How is laminate moisture controlled? | Moisture before lamination can affect interface quality | Storage, handling and baking procedures |
| Can the supplier support custom CAF coupons? | Standard structures may not represent the real product | Coupon design and qualification-planning support |
| How are material substitutions approved? | Equivalent Tg or Dk does not prove equivalent CAF performance | Change-control and customer-approval process |
| How are CAF, SIR and microvia tests differentiated? | Each method addresses a different physical failure path | Test-method and qualification matrix |
| What microsection and traceability evidence is available? | Supports verification and root-cause investigation | Inspection records and material/process traceability |
| How are abnormal resistance trends investigated? | Unstable results require structured failure analysis | Failure-analysis and reporting workflow |
12. Use a Risk Matrix to Select the Required Review Level
A screening matrix can help determine whether a project needs standard DFM, deeper material review or dedicated CAF qualification.
| Risk Level | Geometry | Bias and Environment | Material and Process | Recommended Action |
|---|---|---|---|---|
| Low | Comfortable finished insulation margin | Low bias and controlled environment | Established material and stable process | Standard DFM and material confirmation |
| Medium | Moderately dense conductor geometry | Long-term bias or moderate humidity | Some application data is unavailable | Detailed tolerance and material review |
| High | Small finished spacing or dense hole fields | Continuous bias, high humidity or larger voltage difference | New material, complex lamination or process change | Project-specific coupon and qualification plan |
| Critical | Several high-risk geometric factors combined | Harsh environment and high-consequence application | Limited evidence or uncontrolled substitutions | Design review, material freeze and dedicated qualification |
This matrix is a project-screening tool. It is not a universal CAF acceptance specification and does not replace application-specific engineering analysis.
13. How UltroNiu Supports HDI CAF Risk Review
UltroNiu reviews whether the proposed HDI geometry, material system and reliability requirements can be translated into a controlled fabrication and qualification plan.
Pre-production engineering review
Review stackup architecture, finished spacing, dense via fields, voltage-domain separation, materials and sequential-lamination requirements.
Process and documentation planning
Align material traceability, drilling controls, lamination review, microsection locations and material-substitution restrictions.
Qualification planning
Determine whether CAF, SIR, thermal cycling, IST, microvia testing or other evidence is relevant to the project risk.
Engineering feedback
Return a risk list covering missing geometry information, material concerns, fabrication controls and qualification gaps.
Capability boundary
The review identifies geometry, material and fabrication risks and helps define the required qualification evidence. Project-specific CAF test execution, conditions and acceptance criteria should be confirmed before release.
Review UltroNiu’s HDI PCB manufacturing capabilities or submit the stackup and drill data through the engineering review page.
Frequently Asked Questions
What causes CAF failure in a PCB?
CAF failure occurs when moisture, electrical bias, mobile copper ions and a susceptible internal path exist between conductors. PCB geometry, laminate properties, drilling damage, lamination quality and operating environment can all influence susceptibility.
Why can HDI PCBs be more sensitive to CAF risk?
HDI technology reduces conductor spacing and dielectric margins while increasing interconnect density and fabrication complexity. This does not automatically cause CAF, but it makes dimensional, material and process variation more important.
Is there a universal minimum spacing that prevents CAF?
No. A spacing value cannot be considered universally safe without defining voltage, laminate construction, humidity, temperature, finished manufacturing tolerances, product lifetime and conductor geometry.
How does voltage bias affect CAF formation?
A sustained voltage difference creates the electrical field that drives copper-ion migration. Risk depends on the voltage, dielectric path, moisture exposure and condition of the internal laminate interface.
Does high humidity always cause CAF?
No. Humidity is an enabling condition, but CAF normally also requires electrical bias, a copper source and a susceptible internal migration path.
Can a CAF-resistant laminate eliminate the risk?
No laminate should be treated as completely CAF-proof. Final performance also depends on PCB geometry, material handling, drilling, lamination, electrical bias and the operating environment.
What is the difference between CAF testing and SIR testing?
CAF testing evaluates internal electrochemical migration through the PCB laminate. SIR testing primarily evaluates surface insulation behavior and contamination-related effects. They address different physical failure paths.
Does passing hi-pot testing prove CAF resistance?
No. Hi-pot testing evaluates short-duration dielectric withstand. CAF qualification evaluates internal electrochemical migration under defined environmental stress, electrical bias and exposure time.
When should an HDI PCB use a custom CAF coupon?
A custom coupon should be considered when the design combines small finished spacing, long-term DC bias, demanding humidity exposure, complex lamination, new materials or a high-consequence application.
Can conformal coating prevent internal CAF failure?
Conformal coating mainly protects the board surface. It cannot repair insufficient internal spacing, laminate voids, resin-to-glass separation or drilling damage inside the PCB.
HDI Reliability Review
Review CAF Risk Before the HDI Fabrication Window Is Locked
Send the stackup, drill table, voltage domains, minimum finished spacing, laminate requirements and operating environment. UltroNiu engineers will identify geometry, material and fabrication risks that should be resolved before prototype or production release.