In-Circuit Test and Nail Bed Fixtures

What In-Circuit Test Does

In-circuit test measures each component and each net on a populated board while the board is still in its assembly state. The tester contacts every node through a fixture and applies signals and measurements that isolate one component at a time, so a resistor can be measured without the rest of the circuit influencing the result. The purpose is to find assembly defects, not to prove the design works.

The defects it catches are the ones that assembly produces: missing components, wrong parts, reversed polarity, open joints, solder bridges, and incorrect values. These are the failures that a functional test may or may not notice, because a circuit can often work while containing a wrong or marginal component.

Because the test is electrical and per-component, it gives a specific diagnosis rather than a general failure. That makes repair faster and reduces the risk of a technician replacing a good part while the real fault remains elsewhere on the board.

How a Fixture Is Built

The fixture is a bed of nails: a plate holding spring-loaded probes positioned to match the test points on the board, with a mechanism to press the board onto the probes consistently. Building it requires the board’s fabrication data, the netlist, and a mechanical design that holds the board flat and aligned.

Probe selection is part of the design. Different tips suit different surfaces, and the spring force has to be enough to penetrate any oxide or coating without damaging the pad. Where the board carries tall components, the fixture is built with cut-outs or a taller frame so the board seats fully on the probes.

The fixture also has to survive production. Probes wear, springs lose force, and the alignment drifts as the plate is compressed thousands of times. A fixture that is not maintained will gradually produce false failures, which erode confidence in the test and cause good boards to be reworked.

Test Points and Layout Requirements

The layout determines whether the test is possible. Every net that needs to be measured must have an accessible test point on the side the fixture probes, usually the bottom. The point must be large enough for the probe tip, clear of components that would prevent the board from seating, and spaced from neighbouring points so that probes do not interfere.

Test points also add board area and can affect signal integrity if they are placed on high-speed nets, so the layout has to balance access against performance. Where a net cannot have a dedicated point, a via or a component pad can serve, provided the probe can contact it reliably without damage.

The fixture can only probe one side unless a double-sided design is used, which adds cost and complexity. For dense boards, the layout should keep as many test points as possible on the single probed side, and any net that is inaccessible should be covered by another test method or accepted as untested.

Measurement Techniques

The tester uses a combination of techniques. Resistance measurement with guarding isolates a component by driving the surrounding nodes so that current flows only through the part being measured. Capacitance and inductance measurements identify parts that a resistance check cannot distinguish. Diode and transistor junctions are checked for correct polarity and function.

Powered tests extend the coverage to active devices, applying a supply and checking that a circuit powers up and that key nodes reach expected voltages. These tests need the fixture to carry power and often require the board to be in a known state, so they are more complex to develop than the passive checks.

Boundary scan provides a different route to coverage on digital devices that support it, using the device’s own test logic to observe and drive pins. It reduces the number of physical probes needed but depends on the devices having the capability and on the nets being accessible in the scan chain.

Coverage and Limitations

Coverage is limited by what the fixture can reach and what the test can isolate. A component that is paralleled with another on the same net may be impossible to measure individually without a guarding scheme, and a net with no test point is simply invisible to the test. High-density boards with fine-pitch parts and components on both sides often have coverage gaps that the layout alone cannot close.

Another limitation is that in-circuit test verifies components, not the design. A board can pass every measurement and still fail because a timing margin is inadequate, a thermal problem exists, or the firmware has an issue. That is why ICT is normally combined with functional test and with inspection rather than used alone.

The cost of a fixture and the time to build it make the method best suited to stable, higher-volume products. Where the design changes frequently or the quantity is small, flying probe or a functional test may cover the same defects at lower cost.

Fixture Maintenance and Yield

Once a fixture is in production it needs a maintenance routine. Probe wear should be checked against a schedule, springs tested for force, and the vacuum or clamping mechanism inspected so that the board seats fully every cycle. A probe that has lost its force will produce an intermittent open that looks like a real defect.

The test program should also be kept in step with the product. Changes to the bill of materials, a component tolerance change, or a new supplier can shift the measured values, and if the limits are not updated the test will start failing good boards. Reviewing the false-call rate periodically is a practical way to catch this before it disrupts production.

Finally, the fixture should be stored and handled so that it stays accurate. A fixture that is dropped or stored with the probes loaded will lose alignment, and the resulting failures will be attributed to the boards rather than to the tool.

FAQ

What does in-circuit test catch that a functional test does not? It catches wrong, missing, reversed, and out-of-tolerance components and individual solder defects. A functional test may pass a board that contains those faults if the circuit still works.

Can every net be tested? No. A net with no accessible test point cannot be probed. Coverage depends on the layout and is usually a percentage rather than the full netlist.

Why does my ICT give false failures? Common causes are worn probes, a fixture that no longer seats the board correctly, test limits that no longer match the components, and contamination on the test points.

Is ICT worth the fixture cost? At stable, higher volumes, yes, because the fixture amortises and the test is fast. At prototype or low volume, flying probe or functional test is usually more economical.

Can ICT replace functional test? No. It verifies components and connections, not the operation of the design under real conditions. The two tests find different classes of defect.

Conclusion

In-circuit test is a fast, specific way to find assembly defects on stable products, and its value depends on a layout that provides test access and on a fixture that is maintained properly. Its limits are coverage and cost, both of which should be evaluated against the product’s volume and criticality. Used alongside functional test and inspection, it closes the gap where assembly errors would otherwise reach the customer. For related topics, see our notes on PCBA testing, PCB design and layout, PCB assembly, and quality management for how test coverage is delivered in 2026.