In-Circuit Testing for PCB Assembly

In-circuit testing (ICT) is a production test that checks the components on an assembled board, one node at a time. It makes physical electrical contact with the board and measures values such as resistance and capacitance, catching defects like solder shorts, open connections, missing parts, and components placed with the wrong value or orientation.

ICT answers a narrow but decisive question: was this board assembled the way the design says it should be?

Key takeaways:

• In-circuit testing verifies that an assembled board was built correctly, checking components and connections node by node before the board moves on.
• Access defines the method. A bed-of-nails fixture is fast and built for volume, while a flying probe needs no fixture and fits prototypes and low runs.
• ICT and functional test answer different questions, structural versus behavioral, and work best in sequence rather than as an either/or choice.

What is In-Circuit Testing?

In-circuit testing belongs to the family of structural tests, the checks that verify how a board was built rather than what it does once it runs. An in-circuit tester makes direct contact with predefined test points on a populated PCB assembly (PCBA) and measures the part attached to each node against the values in the design netlist. It reads the value of passive parts such as resistors and capacitors and checks the junctions of diodes and transistors, while active devices and connectors get verified for presence and correct orientation.

If a resistor that should read 10 kΩ comes back open, or two nets that are supposed to stay separate show near-zero resistance between them, the tester catches it and points to the location.

Most of this work happens with the board unpowered, although both are possible. This power-off stage, sometimes run on its own as a manufacturing defect analyzer (MDA), applies low-level signals across the nodes to measure passive components and connections, which makes it a fast and safe first pass on a freshly assembled board. From there, an ICT program can add a powered stage to confirm that active devices respond, and on programmable parts it can load firmware or configuration while the board still sits on the tester.

What sets ICT apart is the precision of its diagnosis. Because every measurement is tied to a specific node, a failure does not simply report that the board is bad. It tells the operator which component or solder joint to rework, which is why ICT has remained a staple of medium- and high-volume manufacturing for decades.

In-Circuit Testing Methods to Access the Board

Before it can measure anything, an in-circuit tester needs a clean electrical path to the nodes on the board. There are two ways to get there. The right one depends mostly on production volume and how often the design changes.

Bed of nails

A bed-of-nails fixture is a custom plate fitted with an array of spring-loaded pins, often called pogo pins, each aligned to a test point on one specific board. It sounds like a fakir's retirement plan, but in test engineering it's the most comfortable thing that ever happened to a circuit board! When the PCBA is pressed onto the fixture, by vacuum or by mechanical actuation, every pin contacts its assigned node at the same time. The tester then sweeps through its measurements across hundreds or thousands of nodes in a single short cycle.

That parallel access is what makes the bed of nails fast enough for high-volume lines. But the cost is rigidity. Each board variant needs its own fixture, which carries an upfront build and a lead time, and the pins wear with use and call for maintenance.

Flying probe

A flying probe tester drops the fixture entirely. A small number of probes mounted on moving arms travel to each test point in sequence, guided by the test program. Nothing is custom-built, so a new board can be on the tester within hours rather than weeks. The trade-off is speed, since probing nodes one after another is slower than contacting them all at once.

That makes the flying probe a good fit for prototypes and low volumes, especially for designs that are still changing, while the bed of nails takes over once volume climbs.

What about Digital and High-Density Boards?

The bed of nails has a hard limit, and modern boards keep running into it. When a design packs in ball grid arrays whose pins sit hidden under the package, or fine-pitch parts too small to land a probe on, there is nowhere left to put a nail. Put components on both sides of the board and the access problem only deepens. Physical contact, the thing every probe-based method depends on, runs out.

For the digital side of the PCB, the answer is boundary scan, better known by the name of its interface, JTAG. Rather than reaching the pins from outside, boundary scan tests from inside the chips. Compliant integrated circuits carry small test cells between their core logic and each external pin, and a short serial test port shifts patterns through those cells to drive and observe every connection in turn. The tester can then check the interconnects between chips, including the solder joints hidden under a Ball Grid Array (BGA), without a single probe touching them. The method is defined by the IEEE 1149.1 standard and has been in service since 1990, a mature answer to a problem that only grows as boards shrink.

Boundary scan does not replace in-circuit testing. It covers a different part of the board. ICT measures the analog and passive parts, along with the power connections that JTAG cannot see, while boundary scan reaches the dense digital interconnects that a fixture cannot. It comes with its own conditions. It works only on devices that include the test logic, and that access has to be designed in from the start, which is one more reason test strategy belongs in the conversation early rather than after the board is already laid out. 

ICT vs FCT: What’s the Difference?

In-circuit testing (ICT) and functional circuit testing (FCT) get mentioned in the same breath so often that they are easy to confuse, yet they answer two different questions. ICT asks whether the board was built correctly. FCT asks whether the board works. One is a structural check on the assembly, the other a behavioral check on the finished product, and neither result tells you what the other would. 

The two methods sit at different points on the line. ICT usually runs first, soon after assembly, because it is fast and screens out structural defects cheaply before any further value is added to the board. FCT runs later, often as the last gate before a unit ships, once it makes sense to spend the longer test time confirming real behavior. Running the structural check first means a board with a solder short gets caught and reworked early, instead of failing a longer functional test downstream and sending an engineer hunting for the cause (and the money they lost).

This is why, at Averna Powered by Spherea, we treat the two not as rival options but as parts of a single test strategy. Rather than route boards through separate, disconnected stations, the more efficient path is often to bring the relevant tests together in one coordinated flow.

When ICT is the Right Choice (and When it’s Not)

ICT is at its best in one specific situation: a stable design produced in enough volume to pay back the cost of its fixture. When those two conditions hold, little else matches it for catching assembly defects quickly and telling you exactly where they are. When they do not hold, ICT is often the wrong tool, and forcing it can cost more than it saves.

What ICT Does Well: Advantages

• It is fast, since the bed of nails contacts every node at once and clears a board in seconds, which is what makes it economical on a busy line.
• It pinpoints faults, returning each failure tied to a specific component or net, so rework is targeted instead of a hunt.
• It delivers high, repeatable structural coverage on a board designed with proper test access.

Where it Runs Out of Road

• The fixture is a custom, one-board tool that carries an upfront cost and a lead time, plus a layout change after it is built can waste much of that investment.
• It depends on test points designed into the board before the layout is frozen, so testability is a decision taken early or not at all.
• It cannot reach what sits under a BGA, which caps coverage on the densest boards.
• A clean pass confirms only that the board was built correctly, not that it works.

What to Use When ICT is Not the Answer

No single method catches every kind of fault, so when ICT does not fit, the question becomes which PCB testing method covers the gap:

• Flying probe replaces the fixture with moving probes, trading speed for flexibility. It suits prototypes and low volumes, or designs still in flux, where a custom fixture would be wasted.
• Automated optical inspection (AOI) uses cameras to catch surface defects, the missing parts and solder bridges a lens can actually see. It is fast and runs on every board, but it cannot see joints hidden beneath a component.
• X-ray inspection (AXI) sees what AOI cannot, imaging the solder joints under BGAs and Quad Flat NoLeads (QFNs) to check for voids and internal faults. It is slower and more specialized, and often unavoidable on boards with hidden joints.
• Boundary scan reaches the dense digital interconnects that probes cannot, as covered above, on devices that carry the test logic.
• Functional testing answers the one question ICT does not, confirming that the powered board behaves as designed before it ships.

Used together rather than in isolation, these methods cover far more than any one of them can alone, which is where a deliberate test strategy earns its keep.

ICT in Production: A Unified Test Strategy

Up to now, we have looked at in-circuit testing one method at a time, but on a real production line the question that matters is how the methods combine. ICT is one part of a broader set of PCB testing approaches, and its value grows when it is planned alongside the others instead of bolted on at the end. That is the approach we take at Averna. Rather than treat in-circuit test as a standalone station, we design it into a coordinated flow built around the product.

One of our ICT Case Studies

A recent project shows what that looks like in practice. A maker of distribution boards for mission-critical electrical panels needed in-circuit test and high-voltage test on the same boards, across a wide range of variants, without the long teardowns that separate stations would force. We combined ICT and HV testing into a single station, with 24 fixtures covering 22 different board variations and two units tested at once, the whole tester stacked vertically in one TM3000 rack. That removed the manual setup and teardown between tests and recovered floor space, while tightening quality control on boards where failure is not an option.

Averna’s ICT Testing

The same logic scales beyond a single station. Platforms such as Averna's UniLine bring ICT together with functional test and optical inspection, through to end-of-line, on one in-line test platform, so a manufacturer running several product variants can test them on the same line instead of maintaining separate equipment for each.

There is a second payoff that is easy to overlook. Every board that passes through ICT leaves a record behind, a log of every measurement it took and, where a board failed, the exact node responsible. Collected across a run, that data turns testing from a simple pass/fail gate into a feedback loop. Recurring failures point back to a drift somewhere on the line, and yield trends can surface a problem before it grows expensive. On boards bound for regulated or mission-critical use, the same records carry the traceability those programs demand.

The common thread is that test strategy is a design decision, not an afterthought. Deciding early what ICT should cover and how it hands off to functional test is what turns a set of individual methods into a coherent way to ship a reliable board.

If you are weighing how in-circuit test should fit into your production line, our automated test solutions team builds these systems around the product and the volume rather than forcing a standard rig onto every board. 

Written by

Max Schultz

Max Schultz is a Technical Sales Engineer and Systems Design expert at Averna, where he works with manufacturers to define and implement advanced test and automation solutions. Starting his career in software development, he has built expertise across project engineering, project management, technical sales, account management, and system architecture, giving him a unique perspective on the challenges of product validation in production. Max specializes in translating complex technical requirements into scalable testing strategies that support quality and long-term growth.

LinkedIn

You may also be interested in…

Watch the video to see how to accelerate PCBA Testing with fully or semi automated solution.