ICT Fixture Design and Test Access

In circuit test checks components and connections on an assembled board by contacting the copper directly, and everything about the test depends on the fixture reaching that copper reliably. An ict fixture is a mechanical assembly that presses hundreds of probes onto defined points at once and keeps them there while the measurements are taken. Its design is a compromise between coverage, mechanical stability and the board space the designer is willing to give up.

What an ICT Fixture Does

The fixture provides electrical contact to every node the test programme wants to measure, carries the stimulus and measurement signals through a wiring interface, and holds the board in a repeatable position so that the results are comparable from unit to unit. It also has to survive thousands of actuations without losing alignment.

Coverage is the measure of how many nodes are reachable. High coverage allows a fast, thorough test that finds a fault in seconds, while low coverage pushes work to later stages where diagnosis is slower and more expensive.

Test Access and Node Selection

Access is decided at layout. Nodes that carry power, ground, critical signals and the junctions between components are the ones worth probing, and the total number of access points sets the fixture complexity. On a dense board the designer cannot probe everything, so the selection is a deliberate exercise in finding the set of nodes that isolates the largest number of faults.

The practical constraint is the space between probes. Probes need a minimum pitch to fit without interfering, so a dense area of the board may only support access on alternating nodes. That decision belongs in the design review rather than in the fixture shop, because moving a test pad afterwards costs a board revision.

<img src="https://www.gopcba.com/wp-content/uploads/2024/09/RF-PCB.jpg" alt="Bed of nails fixture with spring probes aligned to a PCB” />

Test Pads and their Geometry

A test pad should be large enough for the probe to land on reliably, with a defined keepout that stops other components from obstructing the probe travel. A common size is around 1 mm diameter with a 1.3 mm keepout, though smaller pads are usable where the probe is aligned precisely and the board is flat.

Pads should be on a regular grid where possible, be free of solder mask over the copper, and be reachable from the side of the board that faces the fixture. Where a pad must double as a probe target and a normal circuit feature, its shape has to satisfy both requirements. Our notes on test point design cover the geometry in detail.

Probe Types and Pressure

Spring probes come in families that differ in tip shape, travel and spring force. A sharp tip penetrates flux residue and oxide, a rounded tip is gentler on soft platings, and a crown tip works well on via barrels. Probe pressure is the product of spring force and the number of probes, and it accumulates into a substantial load on the board.

Pressure has to be high enough for reliable contact and low enough not to deflect the board into the probes behind it. On a thin board with hundreds of probes, the deflection can be significant, and the answer is support posts under the board rather than an increase in force.

Fixture Construction and Alignment

The probe plate holds the probes in a drilled pattern derived from the board layout, and the alignment is maintained by tooling pins that engage holes in the panel. The plate material must be dimensionally stable, because a small thermal expansion in the plate moves every probe at once.

Alignment is the property that degrades first. Tooling pins wear, plates creep and probes take a set, so a fixture that passes acceptance when new can begin producing false failures after a few thousand cycles. Regular verification against a known good board is what keeps the measurements meaningful.

Test pads on a board underside with silkscreen references

Vacuum and Mechanical Actuation

Vacuum fixtures pull the board down onto the probes using a sealed chamber, which distributes the load evenly and suits large, dense boards. Mechanical fixtures press the board down with a moving platen, which is faster and needs no vacuum supply but applies its load through the platen rather than through the board.

Both have to guarantee that the board cannot shift during the test. Where the board is warped or the panel is thin, a vacuum fixture is more tolerant, while a mechanical press is simpler to maintain and easier to integrate into a high volume line. The choice follows the board rather than the preference of the test engineer.

Maintenance and Probe Life

Probes are consumable. Their spring force falls with use, their tips wear and their barrels collect flux and debris. A maintenance schedule that includes cleaning, force checks on a sample and replacement of worn probes keeps the contact resistance stable and prevents the slow drift that turns into unexplained yield loss.

Cleaning chemistry matters as well, because a probe that is cleaned with the wrong solvent loses its plating. Records of probe replacement intervals, correlated with the false failure rate, give a clear signal about when a fixture needs rebuilding rather than adjusting.

When to Choose Flying Probe Instead

A flying probe machine moves a small number of probes to each node in turn, so it needs no fixture and no dedicated tooling. It is slower per board but ideal for low volume, for prototypes and for boards where access is awkward. Our note on flying probe testing explains where it fits.

Many programmes use both: flying probe for the first article and for small batches, and a fixture for volume once the design is stable. Where a fault is hard to locate, the diagnostic techniques in our article on short circuit inspection apply to either platform.

Process Control and Verification

A short note on the drawing about handling, storage or packaging is often worth more than an extra decimal place on a tolerance.

FAQ

How many test points are enough? Enough to isolate the faults that matter, which usually means most nets rather than all of them. The selection should be reviewed with the test engineer before the layout is frozen.

Can a via be used as a test point? Yes, provided it is not covered by mask and the probe tip suits the barrel. Untented vias on a regular grid are convenient access points.

Why do results drift over time? Almost always contact resistance rather than the board. Probe wear, contamination and alignment drift change the measurement, which is why a known good board is used to verify the fixture regularly.

Leave A Comment