ICT Fixture Probe Selection for Assembled Board Testing

In-circuit test still catches the faults that optical inspection cannot see: a wrong value, a reversed diode, a short under a shield, a connector pin that was never soldered. The inspection methods that cover the rest of the faults are set out in the guide to short circuit inspection methods. The fixture is what makes that possible, and the probes are the part of the fixture that touches the board. Choosing them well is a matter of matching the probe to the pad, the force to the board, and the access to the design that the layout allows.

What an ICT Fixture Has to Contact

The fixture has to reach every node that the test program requires, which in practice means most of the nets on the board. That contact happens on test points, on vias, on component pads and sometimes on connector pins, and each of those surfaces behaves differently under a probe.

A large test point is easy to contact and tolerant of slight misalignment, while a via or a fine pad demands a smaller probe, a lower force and a more accurate fixture. The access that the layout provides therefore sets the limits of the test long before the fixture is built.

Probe Types and Their Uses

The common probe is a spring loaded barrel with a plunger, and the tip geometry is chosen for the surface it will touch. A crown tip cuts through flux and oxide on a solder pad, a conical tip is used where a small contact area is required, and a flat tip spreads the force over a larger area.

Special probes exist for fine pitch work, for high current paths and for coaxial measurements. A high current probe uses a heavier spring and a larger barrel because the resistance of the contact has to stay low, while a coaxial probe carries a shield to the barrel for an impedance controlled measurement.

Probe Force and Spring Selection

Each probe applies a force that depends on its spring, and the total force on the board is the sum of all the probes in contact. A board with a thousand probes and a two newton spring carries a load of two thousand newtons, which is enough to bend a thin panel if the support underneath is not complete.

The force also has to be sufficient to penetrate the surface film on the pad without damaging it. Too little force gives an intermittent contact that appears as a random failure, while too much marks the pad and can crack a small component or a ceramic substrate.

Test Point Design and Access

A test point should be a defined pad of a stated size on a net that will not be disturbed by the probe, and it should be free of solder mask and clear of tall components. The probe needs a vertical approach, so a test point under a tall connector or beside a shield can is effectively inaccessible.

Access on both sides of the board is valuable, because it halves the number of probes per side and reduces the load on the panel. Where the layout allows test points on the bottom only, the fixture has to reach past the assembly side, which limits what can be placed near them.

Fixture Wiring and Signal Integrity

The wiring between the probe and the tester is part of the measurement circuit, and its resistance and capacitance affect the reading. Long wires add resistance to a low value measurement, and unshielded wires pick up noise that appears as a measurement error rather than as a clear failure.

Analog and high frequency measurements are the most sensitive and usually need a short, shielded path from the probe to the instrument. Digital and simple continuity tests are more tolerant, as the comparison with a flying probe tester illustrates, which is why a fixture can sometimes be simplified for a purely digital board.

Probe Wear, Cleaning and Replacement

A probe wears at the tip, where the plating is removed by repeated contact, and its spring loses force over time. Both changes reduce the quality of contact gradually, so the failures they cause appear as a slow rise in the number of intermittent results rather than as a single event.

Probes should be inspected and replaced on a schedule, and the fixture should be cleaned at the same time, because flux and dust on the barrel change the travel of the plunger. The replacement interval is normally set from the number of boards tested rather than from the calendar.

Vacuum Fixtures Versus Mechanical

A vacuum fixture pulls the board down onto the probes with a uniform force, and it is the standard for a large board with a high probe count. A mechanical fixture presses the probes up against the board with clamps or a press, and it is used for small boards and for short runs.

The vacuum version needs a good seal and a flat board, because air leaks at the edges reduce the force available in the middle. The mechanical version needs an even clamp pressure, for the same reason in reverse.

False Failures and Their Causes

A false failure is a test result that says the board is bad when it is not, and it costs time in diagnosis and confidence in the test. Contaminated probes, insufficient force, worn tips and a board that is not seated correctly all produce them.

The pattern of the failures usually identifies the cause: a failure that moves between boards of the same type points at the fixture, one that stays on the same node points at the design or the process, and one that appears only on the first boards of a run points at seating or at cold probes.

Qualification and Maintenance Records

New fixtures should be qualified against a set of known good and known bad boards before they enter production, and the result recorded with the fixture drawing and its probe list. The probe list should state the type, the spring force and the position of every probe, because that is what makes a repair possible without a complete rebuild.

Maintenance records then provide the history of the fixture: probes replaced, cleaning carried out, and any change of tester or program. Where a board fails intermittently, that history is often the fastest route to the cause, and the test structures that support the work are described in the guide to test coupons.

FAQ

How much force should a probe apply? Enough to break through the surface film on the pad and hold contact through vibration, but not so much that the pad or a nearby component is damaged. The figure comes from the probe spring and is normally stated by the fixture designer.

Can a via be used as a test point? It can if it is not covered by mask and if the probe can reach it without touching a component. The contact area is small, so a fine probe and a controlled force are needed.

Why do first boards of a run fail more often? Because the probes and the seating are cold, and because the first boards of a run test a fixture that has been idle. Running a known good board first is a simple way to warm the fixture and confirm the setup.