Fixture Probe: Design Rules and Process Limits

In-circuit test finds assembly defects by touching every accessible node of a board at once. The fixture that does the touching is a block of probes arranged to match the board, and its quality decides whether the test detects real defects or produces a stream of false failures. A fixture probe is a spring loaded contact with a defined travel, force and tip geometry, and each of those properties has to suit the pad it touches. This article explains how fixture design, probe selection and test point design rules fit together.

What In-Circuit Test Measures

An in-circuit tester applies a small current or voltage to a node and reads the result, which allows it to check continuity, detect shorts, measure resistance and verify component values. Because each node is probed individually, the test can isolate a single defective joint or a single wrong part without needing the board to function. That is its main advantage over functional test, which only shows that something is wrong somewhere in the assembly.

The accuracy of that isolation depends on the electrical quality of the contact. A probe that touches a pad through an oxide layer adds resistance in series with the measurement, and a probe that only grazes a pad adds an unstable contact that changes from one insertion to the next. The fixture must therefore make a consistent connection to every node, every cycle, for the life of the product.

Bed of Nails Architecture

A bed of nails fixture holds the board on tooling pins, then closes a probe plate against the bottom side. Each probe is mounted in a receptacle that allows it to compress by a defined travel, and the aggregate force of several hundred probes is what presses the board against its stops. Vacuum fixtures use a gasket and a vacuum chamber to generate that force uniformly, while mechanical fixtures use a press or over-centre clamps.

The mechanical design has to balance two requirements. Force must be enough that every probe reaches its working travel, which means the total load can easily exceed a hundred kilograms, and it must be distributed so the board does not bow between the stops. Boards with thin sections, large cut outs or connectors that cannot be touched need local supports, and those supports are as much a part of the fixture as the probes themselves.

Fixture Probe Selection

Probe selection starts with the pad. A standard spring probe with a crown tip is the default for a round or square test pad, because the serrated tip penetrates flux and oxide and gives a stable contact. A flat tip is used where the surface must not be marked, a sharp tip is used for fine vias, and a radius tip is used on solder joints where a crown would dig in too far.

Spring force and travel come next. The probe must supply enough force to break through contamination but not so much that it dents a small pad or lifts a delicate trace. Working travel is normally about two thirds of the available travel, which leaves margin for board thickness variation and for the change in force as the spring ages. Receptacles should be matched to the probe series so the spring rate is not altered by a mismatched housing.

Test Point Design Rules

Good test points are designed in, not added later. A pad of at least one millimetre diameter with a defined clearance ring around it gives the probe a reliable target, and spacing of at least two millimetres between adjacent points keeps the receptacles from interfering mechanically. Points should be placed on a regular grid where possible, because that reduces fixture cost and makes probes easier to replace.

Accessibility matters as much as size. A test point under a tall component cannot be reached once the part is placed, and a point near a connector body may be blocked by the fixture hardware. Where a node cannot be probed directly, it should be routed to a dedicated point rather than left inaccessible, and the design rules collected in test point design should be applied during layout rather than at fixture build.

Contact Resistance and Measurement Error

Contact resistance is the largest error source in low resistance measurements. A clean probe on clean gold may contribute a few milliohms, while a worn probe on an oxidised pad can contribute hundreds of milliohms, which is enough to hide a trace fault or to fail a good board. The tester normally compensates with a four wire measurement on critical nets, but that only works if there are two probes available for the node.

Where four wire probing is impossible, the practical controls are probe maintenance and pad finish. Keeping pads free of residue before the test, choosing a finish that does not oxidise heavily and replacing probes on a schedule rather than on failure all keep the resistance distribution narrow. Comparing the measured resistance of the same net across many boards is the simplest way to detect a contact problem before it becomes a yield problem.

Probe Wear and Fixture Maintenance

Probes wear in two ways. The tip deforms and loses its ability to penetrate, and the spring loses force as it cycles and collects debris from the pad. Both effects increase contact resistance gradually, so the fixture drifts rather than fails, and the drift is usually attributed to the product until someone measures the probes.

A maintenance schedule should include cleaning the probe tips, checking a sample of probes for spring force and travel, and replacing probes in the highest cycle positions. Recording the number of insertions and the measured resistance on a reference board makes the schedule evidence based. Where a fixture reaches end of life, the alternative approaches compared in flying probe testing may be the better economic choice for low volume.

Fixture Build and First Article

The fixture build should start from the same data the board was manufactured from, not from a physical sample. Using the physical board introduces any error in that one sample, and it becomes permanent in the fixture. Once built, the fixture is verified with a known good board, a board with deliberately introduced opens and shorts, and a stability run that repeats the test many times without a failure.

That stability run is the real acceptance test. A fixture that passes one board and produces intermittent failures after fifty insertions has a contact problem that will appear in production. Reviewing the failure pattern, described in PCBA short circuit inspection, helps separate fixture induced failures from genuine assembly defects before the yield data is trusted.

FAQ

How many test points are practical on one fixture? Several thousand are possible, but force and board stiffness become the limits. Above roughly a thousand probes, vacuum fixturing and a stiffener are usually required.

Should test pads be gold plated? Gold gives the most stable contact resistance, but it is not essential if pads are kept clean and probes are maintained. A solder coated pad works with a crown tip that penetrates the surface.

How often should probes be replaced? Base it on insertions and on measured resistance rather than a fixed calendar. High cycle positions on thick boards with a hard finish wear fastest, and those should be checked first.