How to Test Circuit Boards

Nobody teaches engineers how to test circuit boards at scale. You learn PCB design in school, maybe firmware development, probably some signal integrity. But the part where you make sure thousands of assembled boards actually work before they ship? You figure that out on your own, usually under deadline pressure, usually after something has already gone wrong.

Here is what typically happens: a team designs a board, sends it to a contract manufacturer, and realizes too late that they have no repeatable way to verify each unit. Or they hand-test prototypes for months, then scramble when production volumes make that impossible. Or they skip board-level testing entirely and discover the cost when field returns start coming in.

PCBA testing -- also called PCB assembly testing -- is the process of verifying that assembled circuit boards meet specifications, and it is a different discipline from the design work that created them. This guide covers the major circuit card assembly testing methods, when each makes sense, and how your testing requirements will change as you move from prototype to production.

Why board-level testing matters#

Defects found later cost exponentially more to fix. This is not industry wisdom -- it is basic economics.

A component placed incorrectly during assembly costs pennies to fix on the production line. The same defect, discovered during system integration, might require hours of debug time. Found by a customer in the field? You are looking at warranty costs, returns, support calls, and damaged reputation.

The goal is not testing for testing's sake. It is catching manufacturing defects before they compound into expensive problems.

PCBA testing methods#

Several distinct approaches exist for automated PCB testing. Each targets different defect types and suits different situations. The right testing strategy usually combines two or more of these methods rather than relying on any single one.

Automated optical inspection (AOI)#

AOI systems use cameras to visually inspect boards after assembly. They catch visible defects: missing components, misaligned parts, solder bridges, tombstoned passives.

AOI inspects every board without slowing the production line and requires no test fixture. It catches visible defects reliably and works on any board topology. But it cannot verify component values, detect hidden solder joint defects like cold joints or head-in-pillow, or confirm that a board actually functions. False positive rates vary by setup quality.

AOI works best as a first-pass screen, catching obvious manufacturing issues before boards proceed to electrical testing.

In-circuit testing (ICT)#

ICT uses a bed-of-nails tester — a fixture with spring-loaded probes contacting test points across the board — to electrically verify component presence, values, and connections. This is component-level verification -- not just "board works," but confirmation that each resistor, capacitor, and IC is the right part, in the right place, at the right value.

ICT catches wrong component values, missing components, reversed polarity, solder opens and shorts, and incorrect component orientation. Cycle times are fast enough for production volume, measurements are quantitative (actual values, not just pass/fail), and defects isolate to specific components.

The tradeoff: ICT requires a dedicated PCBA test fixture, which means cost and lead time. Your PCB design needs accessible test points. The fixture must match the board revision. And ICT alone does not verify full system functionality.

For detailed information on ICT implementation, see our In-Circuit Testing guide.

Functional testing (FCT)#

PCBA functional testing verifies that the assembled board performs its intended function. The board is powered and exercised through its normal operating modes -- the closest thing to asking "does this actually work?" before shipping.

For hardware products at production volume, a functional tester typically uses a bed-of-nails test fixture to make repeatable connections between the board and test instrumentation. The fixture handles the mechanical interface: power, ground, test signals, and programming connections all made in one press. This is where fixtures and functional testing intersect -- the fixture provides reliable physical access, while the test software validates board behavior.

Functional testing catches system-level defects ICT misses, validates firmware and software interaction, and can include environmental stress like temperature or voltage variation. The limitation is that a binary pass/fail often provides poor defect localization. When a board fails functional test, you know something is wrong but not always what. Cycle times run longer than ICT, and test development can be complex.

For most production scenarios, functional testing complements rather than replaces component-level testing. See our Functional Testing guide for implementation details.

Automated X-ray inspection (AXI)#

AXI uses computer vision on X-ray images rather than optical images, validating solder connections including those hidden under components. This is particularly effective for ball grid array (BGA) packages where optical inspection cannot see the solder joints.

AXI works well for BGA and CSP package solder joint verification, hidden solder joints under RF shields or heat sinks, void detection in solder paste, and through-hole solder fill verification. Equipment cost is higher than AOI, throughput is slower, complex cases require operator interpretation, and it cannot verify component values or functionality.

AXI is often paired with AOI -- AOI catches surface-visible defects quickly while AXI verifies hidden connections on critical assemblies.

Boundary scan (JTAG)#

Boundary scan uses built-in test infrastructure in digital ICs (defined by IEEE 1149.1) to test connections between chips without physical probe access. It is increasingly important as board density increases and probe access decreases.

Boundary scan is well suited for BGA devices with no probe access to balls, dense digital boards with limited test points, and verifying connections on high-pin-count processors. It only works with boundary scan-compliant devices, does not test analog components, requires JTAG chain design consideration, and has programming and setup complexity.

Flying probe#

Flying probe testers use movable probes rather than a fixed fixture, trading speed for flexibility.

Flying probe makes sense for low volume production where fixture cost is not justified, prototype and NPI builds, mix of board variants on the same line, and quick-turn production. When volume exceeds roughly 500-1,000 units, cycle time requirements demand parallel test, or per-unit test cost becomes significant, it is time to move to fixture-based testing.

See our Flying Probe guide for a detailed comparison with fixture-based ICT.

When to use each method#

Most production testing combines multiple methods. The right mix depends on your specific situation.

By production volume#

Prototype/NPI (1-100 units): AOI plus functional testing is often sufficient. Flying probe works if you need electrical verification. Fixture investment usually does not make sense at this point.

Low volume (100-1,000 units): AOI plus functional as a baseline. Flying probe or a simple fixture depending on board complexity. Evaluate fixture ROI based on expected total volume.

Production volume (1,000+ units): AOI as first-pass screen. ICT with a dedicated fixture for component verification. Functional testing for system validation. Consider boundary scan for BGA-heavy designs.

By board complexity#

Simple boards (few components, low density): AOI and functional may be sufficient. ICT adds value when component values are critical.

Medium complexity (mixed signal, moderate density): ICT strongly recommended for component verification. Functional testing validates overall operation. AOI catches visible assembly defects.

High complexity (BGAs, high density, mixed signal): Multiple methods required. Boundary scan for BGA interconnect verification. ICT where test points are accessible. Functional testing for system validation. AOI for visible defect screening.

How testing requirements change with volume#

Testing strategy is not static. As you scale from prototype to production, requirements shift. For detailed planning guidance on this transition, see our guide on testing from prototype to production.

Prototype stage#

At prototype, you are debugging the design as much as verifying assembly quality. Testing tends to be manual, exploratory, and focused on "does it work at all?"

Typical approach: manual bring-up and bench testing, basic functional verification, debug-as-you-go mentality.

What often goes wrong: no documented test procedure, defects found through system-level symptoms, poor traceability of what has been verified.

Pre-production stage#

As you move toward pilot builds and contract manufacturer handoff, testing needs to become repeatable and documented.

Required changes: written test procedures that others can execute, pass/fail criteria (not just "engineer says it's good"), test data logging for trend analysis, and fixture investment decisions.

This is where many teams realize their prototype testing approach will not scale. Building internal fixtures or relying on functional testing alone often creates bottlenecks.

Production stage#

At production volume, testing must be fast, repeatable, and cost-effective per unit.

Key requirements: cycle time matters (seconds, not minutes), test equipment utilization becomes significant, defect data enables process improvement, and fixture durability affects ongoing costs.

Common mistakes and how to avoid them#

Years of working with engineering teams have revealed consistent patterns in testing mistakes.

Mistake 1: No test points in the design#

If your PCB layout does not include accessible test points, your testing options narrow significantly. Adding test points post-layout is expensive.

Prevention: Include test strategy discussion in design reviews. Specify minimum test point size (0.9mm+ diameter) and spacing (1.9mm+ minimum) for reliable probing.

See our Design for Test guide for detailed design rules.

Mistake 2: Functional testing only#

When a board fails functional test, you know something is wrong but often not what. That's the problem with relying solely on functional testing — it seems efficient (if it works, ship it), but defect isolation suffers.

Use ICT or component-level testing to identify specific defects, then functional testing to verify system operation.

Mistake 3: Over-testing prototypes#

Production-grade testing on prototype builds wastes money. Your design will change; your fixture becomes obsolete.

Better approach: Match test investment to production expectations. Use flying probe or simple fixtures for prototype and NPI. Invest in production fixtures when design is stable.

Mistake 4: Fixture as afterthought#

Treating fixture procurement as a last-minute item creates schedule problems. Fixtures require design, manufacture, and validation.

“

Fixture needs to be equal or higher quality than the product itself. Huge underestimation of how important they are.

”

— Test Engineering Manager, Consumer Electronics

Better approach: Include fixture lead time in your production timeline. Start fixture discussions before design freeze.

Mistake 5: Ignoring test data#

Test data reveals process problems before they become yield issues — but only if someone looks at it. Running tests without analyzing results means repeating the same problems at higher volume.

Log test results, trend key parameters, and set statistical limits for early warning.

Building your test strategy#

What defects are you trying to catch? Manufacturing defects (ICT), design defects (functional testing), or both?

What is your expected volume? This determines fixture investment ROI.

Where will testing happen? In-house capability versus contract manufacturer requirements differ.

What test data do you need? For process improvement, regulatory compliance, or customer requirements?

Next steps#

Based on where you are in your product development:

If you are in early design:

• Review our Design for Test guide
• Ensure test points are in your layout
• Consider test strategy before layout freeze

If you are approaching production:

• Evaluate build vs. buy for test fixtures
• Understand testing method tradeoffs
• Plan fixture lead time into your schedule

If you are scaling to volume:

• Development fixtures cover prototype and NPI builds; higher-durability options handle pre-production volume and tighter pitch requirements. See fixture options to match your stage.
• Consider the prototype to production transition