CAF Failure Mechanisms
Quality & Inspection Standards
PRODUCTS CENTER
Get Custom Quote
PRODUCTS CENTER
CAF Failure Mechanisms – IPC‑6012 Limits, Test Methods & Design Prevention
Conductive Anodic Filament (CAF) is one of the most insidious field failure mechanisms in high‑density PCBs. It does not appear during electrical test. It does not show up in thermal stress. It grows slowly – over months or years – along glass fibers inside the laminate, under humidity and voltage bias. We have seen telecom backplanes fail after 18 months of field operation because CAF grew between adjacent PTHs spaced only 0.4mm apart. The board had passed all qualification tests: electrical continuity, insulation resistance, thermal stress, even ROSE. CAF is a materials and design problem, not a cleanliness one. Per IPC‑6012E §3.7.3, CAF resistance testing (IPC‑TM‑650 2.6.25) is required for Class 3 products when specified. This guide covers CAF formation physics, critical parameters (spacing, voltage, humidity, material), test methods, and design rules to prevent CAF in high‑reliability PCBs.
Table of Contents
- CAF Physics – The Four Necessary Conditions
- Critical Parameters: Spacing, Voltage, Humidity, Glass Fiber Orientation
- CAF Test Method – IPC‑TM‑650 2.6.25
- Material Influence – Laminate Type, Resin Chemistry & Glass Treatment
- Design Rules to Prevent CAF – Minimum Spacing & Layout Practices
- Cost vs Reliability – When to Specify CAF‑Resistant Materials
- CAF Prevention Checklist
- FAQ
1. CAF Physics – The Four Necessary Conditions
CAF is an electrochemical failure mechanism. It requires all four of the following to occur simultaneously:
- 1. A conductive path (glass fiber bundle): The filament grows along the interface between the glass fiber and the resin. Epoxy‑glass laminates (FR‑4) are susceptible; non‑woven aramid or PTFE are less so.
- 2. An ionic contaminant (source of mobile ions): Chloride (Cl⁻), bromide (Br⁻), or copper ions (Cu²⁺). Often from flux residues, solder mask developer residues, or plating chemistry.
- 3. Moisture (humid environment): Typically >70% RH. The moisture allows ion migration. CAF is rare in dry environments.
- 4. A voltage bias (potential difference): Typically >10V DC, but has been observed at 5V in fine‑spacing designs. The anode (+) loses copper ions; the cathode (−) attracts them.
The filament grows from the anode toward the cathode. Under a microscope, it appears as a dark, dendritic structure along the glass fiber. Once the filament bridges the two conductors, a short circuit (or low‑resistance path) occurs. The failure is often intermittent – resistance may be high at room temperature but drop when humidity increases or temperature rises.
In a 48V power distribution board, CAF grew between adjacent PTHs (0.6mm spacing) after 24 months. The filament was only visible at 100x. The board passed ICT (in‑circuit test) because the short was intermittent – only when the board was warm and humid. The root cause was a combination of insufficient spacing and a halogenated flame retardant that released bromide over time.
2. Critical Parameters: Spacing, Voltage, Humidity, Glass Fiber Orientation
The time‑to‑failure (TTF) for CAF is governed by several variables. The table below shows typical acceleration factors (industry data, not IPC).
| Parameter | Safe Zone | CAF Risk Zone | Acceleration Factor |
|---|
Table 1: CAF risk factors. Spacing is the most controllable design parameter.
In one study, boards with 0.3mm spacing failed CAF testing (500 hours, 85°C/85%RH, 50V) at 100%, while identical boards with 0.5mm spacing had 0% failures. The relationship is highly non‑linear. For high‑reliability products (telecom, automotive, aerospace), we recommend minimum conductor spacing of 0.5mm for any biased net.
3. CAF Test Method – IPC‑TM‑650 2.6.25
IPC‑TM‑650 2.6.25 specifies the CAF resistance test. Key parameters:
- Test vehicle: A specially designed coupon with two or more adjacent conductor patterns (e.g., PTHs or vias) at defined spacings (0.4mm, 0.5mm, etc.).
- Temperature: 85°C ± 2°C.
- Humidity: 85% ± 3% RH.
- Bias voltage: Typically 50V or 100V DC, applied continuously.
- Duration: 500 hours minimum (some specifications require 1000 hours).
- Measurement: Insulation resistance measured at intervals (e.g., 24, 48, 96, 168, 500 hours). A drop below 1×10⁸ Ω (or 1×10⁷ Ω) indicates CAF.
- Post‑test analysis: Microsection at 200‑500x to visually identify filaments along glass fibers.
The test is destructive and expensive ($1,000‑$3,000 per coupon). It is typically run on first article qualification, not on every lot. Many OEMs require CAF testing for Class 3 products with conductor spacing <0.5mm.
In one qualification program, a high‑Tg FR‑4 (170°C) passed 500 hours of CAF testing at 0.5mm spacing but failed at 0.4mm spacing after 350 hours. The supplier changed to a "CAF‑resistant" laminate (special resin chemistry) – the same design passed 1000 hours. The material cost increase was 12%, well justified for a medical device.
4. Material Influence – Laminate Type, Resin Chemistry & Glass Treatment
Not all laminates are equal in CAF resistance. The key material factors:
- Resin chemistry: Standard FR‑4 uses brominated epoxy, which can release bromide ions over time (contributing to CAF). Halogen‑free laminates (phosphorus‑based) may have better CAF resistance but lower Td. Some specialty resins (e.g., certain high‑Tg epoxies, polyimide, BT) are inherently more CAF‑resistant.
- Glass fiber treatment: The coupling agent (silane) that bonds glass to resin affects CAF. Poor adhesion allows moisture wicking. "CAF‑resistant" laminates use optimized silane chemistry.
- Glass weave style: Open weaves (e.g., 106, 1080) have larger channels for CAF growth. Dense weaves (e.g., 2116, 7628) provide less path continuity.
IPC‑4101 slash sheets have CAF resistance designations. For example, /124 (high‑Tg FR‑4) is available with "CAF‑resistant" option. Ask your laminate supplier for CAF test data – most have standard data for common spacings (0.4mm, 0.5mm, 0.6mm).
In a server backplane design, the original FR‑4 (standard) failed CAF testing at 400 hours (0.45mm spacing). Switching to a halogen‑free, CAF‑optimized laminate with the same Tg extended TTF to >2000 hours. The cost premium was 15% – negligible compared to the cost of a field recall.
5. Design Rules to Prevent CAF – Minimum Spacing & Layout Practices
The most effective CAF prevention is design – not material selection. Based on industry experience and CAF test data:
- Minimum edge‑to‑edge spacing for biased nets: ≥0.5mm (20mil) for voltages ≤50V. For voltages >50V, add 0.1mm per 10V. For 100V, minimum 1.0mm spacing.
- Route bias‑critical nets away from glass fiber direction: Where possible, route traces perpendicular to the glass weave to avoid long parallel paths.
- Do not place PTHs in a straight line parallel to glass fibers: Staggered via arrays are safer.
- For high‑voltage (>100V), add a routed slot between opposing nets: This physically blocks the CAF path.
- Keep the same net away from opposite polarity nets: CAF requires a voltage gradient. Equal potential nets do not cause CAF.
A power supply design with 200V bias between adjacent traces (spacing 0.8mm) failed CAF after 6 months. Adding a 1mm routed slot between the traces completely eliminated the problem. The slot cost $0.10 per board – far cheaper than a recall.
6. Cost vs Reliability – When to Specify CAF‑Resistant Materials
The cost of CAF prevention increases as spacing shrinks. Tradeoffs:
- ≥0.6mm spacing: Standard FR‑4 is usually sufficient. No special material needed. Cost baseline.
- 0.5mm spacing: Standard FR‑4 may be borderline, especially at high voltage (>48V). Recommend CAF‑optimized FR‑4 (+10‑15%).
- 0.4mm spacing: Standard FR‑4 will likely fail. Require CAF‑resistant laminate (+15‑25%). Some designs can increase spacing to 0.5mm at minimal layout cost – always try that first.
- 0.3mm spacing: Extremely high CAF risk. Even CAF‑resistant materials may fail. Consider redesign to increase spacing or add slots.
In a high‑density automotive board, the layout had 0.35mm spacing between a 48V net and ground. CAF testing failed after 300 hours. Increasing spacing to 0.5mm required moving two traces – a layout change that added 0% material cost and 4 hours of engineering time. The board passed 2000 hours. Always try spacing first before spending money on exotic materials.
CAF Prevention Checklist (IPC‑Anchored)
- Identify all biased nets. Any net with voltage difference >5V is a CAF risk.
- Calculate minimum edge‑to‑edge spacing. Use 0.5mm as baseline for ≤50V. Increase for higher voltage.
- If spacing <0.5mm cannot be avoided, specify CAF‑resistant laminate. Request CAF test data from supplier.
- Route traces perpendicular to glass weave. Avoid long parallel runs.
- Stagger PTHs, do not align them in straight lines.
- For critical high‑voltage nets, add routed slots between opposing polarities.
- Require CAF testing per IPC‑TM‑650 2.6.25 on first article for Class 3 or spacing ≤0.5mm. Specify test duration (500 or 1000 hours) and pass/fail resistance (typically ≥1×10⁸ Ω).
7. FAQ
Q: What is the difference between CAF and dendritic growth?
A: Dendrites grow on the surface of the PCB (solder mask) or through air gaps. CAF grows inside the laminate, along glass fibers. CAF is much harder to detect because it is not visible on the surface. It requires microsection to identify.
Q: Does IPC‑6012E require CAF testing for Class 3?
A: No – IPC‑6012E §3.7.3 says CAF testing is required "when specified in the procurement document." It is not mandatory for all Class 3. However, many aerospace, medical, and automotive customers require it.
Q: Can cleaning reduce CAF risk?
A: Partially. Removing ionic residues (chloride, bromide) from the surface reduces the ion source. However, CAF can also grow using ions already present in the laminate (e.g., from the flame retardant). Cleaning helps but does not eliminate CAF in contaminated laminates.
Q: What is the minimum spacing that guarantees no CAF?
A: There is no guarantee – CAF has been reported at 1.0mm spacing under extreme conditions (high voltage, high humidity, aggressive contamination). For practical designs, 0.5mm spacing with standard FR‑4 and good cleanliness is considered safe for <50V. For higher reliability, use 0.6mm.
Continue Reading – Cleanliness & Contamination Series
Concerned about CAF in your high‑density design? Our engineers will review your layout, recommend minimum spacing, and help specify CAF‑resistant materials or CAF test plans per IPC‑6012 and IPC‑TM‑650.
Request CAF Risk Assessment →References: IPC‑TM‑650 2.6.25; IPC‑6012E §3.7.3; IPC‑2221B §9.2; IPC‑4101 (laminate CAF data). IPC‑TM‑650 Official Method.
Get Custom Quote
PRODUCTS CENTER


