We have been working in the PCB industry for many years. Over this time, we have encountered numerous failure cases that could have been avoided, as well as some that took even us by surprise. In this article, we will discuss the most common causes of PCB failure, how to detect them, and truly effective preventive measures for actual production.

First, a Quick Answer

PCB failures most commonly come down to one of these root causes: thermal stress during or after manufacturing, moisture ingress, design errors that weren’t caught before production, manufacturing defects from process control gaps, and electrical overstress in the field. We’ll break each one down in detail below.

Thermal Stress — The Quiet Killer

This is probably the most underappreciated cause of PCB failure, especially in high-reliability applications.

Every material on a PCB — the substrate, copper traces, solder mask, component leads — expands and contracts at a different rate when temperature changes. That rate is called the Coefficient of Thermal Expansion (CTE). When those rates don’t match up, you get mechanical stress at the interface. Do that enough times across a temperature cycle, and something eventually gives.

What does “giving” look like? Usually it’s a via barrel crack. The via wall, which is a thin cylinder of copper connecting layers, develops a fracture. The board might work fine for months, then start exhibiting intermittent failures that are nearly impossible to reproduce on a test bench at room temperature.

We’ve seen this happen most frequently in:

• Boards that go through many reflow cycles (rework is a big one)
• Automotive or industrial applications with wide operating temperature ranges
• Multilayer boards with mixed-material stackups where CTE mismatch wasn’t accounted for in design

The fix isn’t always obvious. Sometimes it’s material selection — moving to a higher Tg laminate or choosing a substrate with a CTE closer to copper. Sometimes it’s design — adding thermal relief to heavy ground pours, or adjusting via aspect ratios. And sometimes it’s process — making sure reflow profiles are dialed in correctly and boards aren’t being heat-stressed unnecessarily during assembly.

Moisture and Humidity

This one catches people off guard because the damage is often invisible until it’s too late.

PCB laminates, especially standard FR-4, are hygroscopic — they absorb moisture from the air. That absorbed moisture becomes a problem in several scenarios:

During assembly: If a board has been sitting in a humid warehouse and then gets run through a reflow oven, the trapped moisture vaporizes rapidly. The result can be delamination, measling (those white spots inside the laminate), or in severe cases, a phenomenon called “popcorning” in moisture-sensitive components.

In the field: Over time, moisture can facilitate a process called Conductive Anodic Filament (CAF) formation — essentially, copper ions migrate through the laminate between conductors under bias, eventually creating a conductive pathway where there shouldn’t be one. CAF failures are rare but catastrophic when they happen, and they’re extremely difficult to diagnose without cross-section analysis.

Corrosion: Exposed copper, pads that haven’t been properly finished, or damaged solder mask all become corrosion sites in humid environments. We see this most often on boards that were stored improperly or shipped without adequate moisture barrier packaging.

The practical prevention here is almost annoyingly simple: store boards in sealed, desiccated packaging, follow IPC moisture sensitivity level (MSL) guidelines for components, and bake boards before assembly if they’ve been exposed to humidity for extended periods. The fact that these failures still happen frequently tells you that storage practices are not taken seriously enough across the industry.

Manufacturing Defects — Not All Are Created Equal

Let’s be honest: no manufacturing process is perfect. The question is whether your manufacturer has the process control and inspection systems to catch defects before they reach you.

The most common manufacturing-related failures we see are:

Cold solder joints — These form when solder doesn’t fully wet to the pad or lead, often due to insufficient heat, contamination, or incorrect flux. They look fine visually but have poor mechanical and electrical integrity. They tend to fail under vibration or thermal cycling.

Bridging and shorts — Excess solder creating unintended connections between adjacent pads or traces. More common with fine-pitch components and increasingly relevant as designs push toward tighter clearances.

Insufficient copper coverage in vias — Electroplating is a controlled process, but variations in plating chemistry, current density, and board geometry can leave some vias with thinner copper walls than specified. These vias may pass initial testing but fail under thermal stress over time.

Lifted pads — Usually caused by excessive heat during rework, but can also happen during original assembly if the pad adhesion to the substrate is compromised by laminate quality issues.

The IPC standards — particularly IPC-A-600 for bare boards and IPC-A-610 for assemblies — define acceptance criteria for all of these. If your manufacturer can’t tell you which IPC class they’re producing to, that’s a concern.

Design Errors (The Ones Nobody Wants to Talk About)

A lot of the failures we see in the field were designed in — they just didn’t show up until the board was in a real environment.

Some of the most common design-driven failures:

Trace width miscalculation: A trace that’s too narrow for its current load will heat up. Sustained heat degrades the copper, the solder mask above it, and eventually causes either a trace open or a localized fire risk. Online trace width calculators are useful but don’t account for everything — nearby heat sources, copper pour fills, or operating in an enclosure change the thermal picture significantly.

Inadequate thermal management: Power components that dissipate heat need thermal vias, copper pours, or other pathways to move that heat away. Boards where a power component sits on a small, isolated pad with no thermal connection to a ground plane tend to fail early — the component gets hot, cycles thermally, and the solder joint fatigues.

Spacing violations near high-voltage nodes: We’ve seen creepage and clearance distances that technically met the design rules but weren’t appropriate for the operating environment. Boards that operate in humid or contaminated atmospheres need more spacing than boards in a sealed, controlled enclosure.

No design for manufacturability (DFM) review: Features like trace-to-edge proximity, drill-to-copper clearances, and solder mask slivers all affect whether a board can actually be manufactured reliably. Catching these before fabrication starts saves everyone time and money.

A good DFM review from your PCB manufacturer before you send the files for production isn’t a bureaucratic formality — it’s one of the highest-value activities in the product development process.

Electrostatic Discharge (ESD)

ESD damage is notoriously tricky because most of it is latent. A component that gets zapped by a 500V discharge during handling might not fail immediately. It might pass all your incoming inspection tests. It might work for six months. And then it fails in the field, and nobody connects the dots back to a moment at a workbench where someone didn’t use a wrist strap.

Board-level ESD protection — TVS diodes, ESD suppression arrays, proper ground plane design — helps protect against ESD events in the field. But handling-related ESD damage happens before any of that matters.

The fundamentals are straightforward: ESD-safe workbenches, wrist straps with continuity testing, ESD bags for storage and shipping, and training for anyone who handles bare or assembled boards. It’s not glamorous, but it’s effective.

Mechanical Stress

PCBs are rigid, but they’re not indestructible. Boards that are flexed during assembly (from being pressed into tight enclosures, for example), subjected to vibration in service, or impacted during shipping can develop cracks — either in the board substrate itself, in solder joints, or in component bodies.

BGA and chip-scale packages are particularly vulnerable to mechanical stress. The solder balls are small, numerous, and in some areas have very little solder to absorb mechanical strain. A board that gets flexed during assembly can crack solder joints under BGAs without any visible sign of damage — and those cracks may not cause an immediate failure.

If your application involves vibration (automotive, industrial, aviation), board-level mechanical simulation during design, plus appropriate mounting, stiffeners, and vibration isolation, is worth the investment. We’ve seen boards designed for consumer electronics drop catastrophically in high-vibration environments because the mechanical loading was never considered in the design phase.

Chemical Contamination

Flux residue is the most common culprit here. Certain flux chemistries, if not cleaned after assembly, leave residues that are hygroscopic and mildly acidic. Over time, those residues can corrode conductors, degrade surface finishes, and create leakage paths between conductors.

No-clean fluxes are designed to leave benign residues — but “no-clean” doesn’t mean “no residue.” In high-reliability applications, or anywhere the board will be in a humid environment, cleaning is still often the right choice regardless of the flux type.

We’ve also seen contamination from handling — skin oils deposited on bare copper before final finish processing — and from inadequate rinse cycles after chemical processing steps. These are process discipline issues, and they’re the kind of thing that a proper quality management system catches.

Electrical Overstress (EOS)

EOS is different from ESD, though they’re often confused. ESD is a fast, high-voltage, low-energy event. EOS is a slower, lower-voltage, higher-energy event — typically from a power supply transient, a load dump, or an incorrect voltage being applied to a board.

EOS damage tends to be more dramatic and easier to identify: burned traces, cracked components, discolored substrate. But the damage it causes can be just as thorough.

Protection comes from a combination of proper power supply sequencing, input filtering, transient suppression, and — at the system level — understanding what fault conditions the board might realistically see in the field.

Via and Plating Failures

We’ve touched on this under thermal stress and manufacturing, but it deserves its own section because it’s so common in multilayer boards.

The via barrel — the copper cylinder that connects different layers — is stressed every time the board temperature changes. The copper expands and contracts axially, and over thousands of cycles, fatigue cracks can form, typically at the knee of the via (where the barrel meets the pad on an outer layer).

Aspect ratio matters a lot here. A via that’s very deep relative to its diameter is harder to plate uniformly and more prone to barrel cracking. For high-reliability boards, keeping via aspect ratios within conservative limits, using microvias in HDI designs where appropriate, and specifying tighter plating thickness requirements all reduce the risk.

Counterfeit and Poor-Quality Components

This is a real problem, and it’s gotten worse over time. Counterfeit components — either completely fake parts or legitimate parts that have been remarked to misrepresent their specifications — find their way into supply chains through grey-market distributors.

The failure modes are varied: a component that’s rated for 85°C ambient when the application runs at 105°C, a MOSFET with worse-than-specified RDS(on) that runs hot under load, a capacitor that’s been remarked with a higher voltage rating than it can actually handle. These parts may pass functional testing but fail prematurely in the field.

Buying from authorized distributors, running incoming inspection on critical components, and having a supply chain quality process are the main defenses. If a price seems too good to be true, it usually is.

How Do You Actually Detect PCB Failure?

Knowing something failed is rarely enough — you need to know why it failed.

Automated Optical Inspection (AOI) catches surface-level assembly defects: missing components, solder bridges, lifted leads, polarity errors. It’s fast and effective for catching manufacturing defects, but it can’t see inside solder joints or beneath BGAs.

X-ray inspection is essential for area-array packages like BGAs. It gives you a look at solder joint morphology — voids, head-in-pillow defects, bridging — that’s invisible to optical inspection.

Thermal imaging (infrared) can locate hot spots on a powered board, which often corresponds to a shorted component, a trace with too much resistance, or a thermal management problem. It’s a fast, non-destructive first pass for field failures.

Flying probe and in-circuit testing verify electrical continuity and component values. They catch opens, shorts, and wrong-value components.

Cross-section analysis (metallographic sectioning) is the definitive diagnostic tool for understanding via integrity, plating thickness, laminate quality, and solder joint microstructure. It’s destructive — you sacrifice the board — but when you need to know exactly what happened at a microstructure level, there’s no substitute.

Preventing PCB Failure: What Actually Works

There’s a lot of generic advice out there — “follow IPC standards,” “do DFM reviews” — and while that’s all true, here’s the more specific version:

At the design stage: Get a DFM review from your manufacturer before you finalize the files. Ask specifically about via aspect ratios, thermal management, trace widths, and spacing. If your manufacturer just says “looks fine,” find a different manufacturer.

At the component sourcing stage: Use authorized distributors for critical components. Define a short list of acceptable alternates and qualify them properly rather than approving random substitutions.

During manufacturing: Understand what your manufacturer’s process controls look like. What SPC data do they collect on their plating lines? What’s their AOI coverage? What’s their first-pass yield, and what happens to boards that fail?

For high-reliability applications: Specify reliability testing — HALT (Highly Accelerated Life Testing) or HASS (Highly Accelerated Stress Screening) — to stress-test the design and catch latent defects before boards go into the field. It’s more expensive upfront, but it’s far less expensive than a field failure.

FAQ

What is the most common cause of PCB failure? In our experience, thermal stress and manufacturing defects together account for the majority of PCB failures. Thermal stress failures often appear later in the field as intermittent issues, while manufacturing defects typically show up earlier in the product lifecycle.

How long does a PCB last? A well-designed and well-manufactured PCB in appropriate operating conditions can last 20years or more. The actual lifespan depends heavily on thermal cycling frequency, environmental conditions, component quality, and whether the board was designed with appropriate derating and protection. Boards in harsh environments (automotive, industrial) often have specific reliability targets defined in cycles or hours rather than calendar time.

Can a failed PCB be repaired? It depends on the failure mode. Many assembly-level defects — cold solder joints, lifted components, solder bridges — can be repaired by a competent rework station. Laminate damage, via barrel cracking deep within a multilayer board, or widespread corrosion typically can’t be economically repaired. For mission-critical applications, we generally recommend replacing rather than reworking a failed board once the root cause is understood.

What is PCB delamination? Delamination is the physical separation of the laminate layers within a PCB. It’s caused by moisture vaporizing rapidly under heat (during assembly or rework), by thermal cycling beyond the laminate’s capability, or by poor laminate quality. Delaminated boards typically show a visible blister or separation and almost always need to be scrapped.

How do you perform a PCB failure analysis? A systematic failure analysis starts with understanding the symptoms and operating history, followed by non-destructive inspection (visual, X-ray, thermal imaging), then electrical testing to characterize the failure mode, and finally destructive analysis (cross-section, SEM) if needed to identify the root cause. The goal isn’t just to identify the failed component — it’s to understand the mechanism so the failure doesn’t happen again.

A Final Thought

PCB failure is rarely mysterious once you’ve seen enough of it. The same failure modes show up repeatedly, and they almost always trace back to a gap somewhere in the design-manufacturing-supply chain process.

The boards that don’t fail are the ones where someone — a designer, a manufacturing engineer, a quality manager — thought carefully about what could go wrong and made decisions to prevent it. That’s not luck. It’s engineering discipline.

If you’re evaluating PCB manufacturers and want to understand how we approach quality and process control, we’re happy to walk you through it. Feel free to reach out for a DFM review on your next project — it’s free, and it’s often the most useful conversation we have with new customers.

The post What Causes PCB Failure? A PCB Manufacturer’s Honest Breakdown first appeared on thindry pcb manufacturer.