Functional Test Fixture Design for PCBA

A functional test fixture connects the product to the test system, and every connection it makes has to be reliable, repeatable and harmless. The fixture is designed from the test requirement backwards, which is the opposite direction from the way a board is usually laid out.

Starting From the Test Requirement

The first question is what the test has to prove and how many boards it has to prove it on. A fixture for a hundred prototypes is built differently from one that runs a million units, and the difference is in the wear parts rather than in the electronics.

Where the requirement is not written down, the fixture designer supplies a default and that default is chosen for the drawing rather than for the production rate. Our test point notes describe the board side of the same decision.

Probe Selection and Wear

A spring probe has a travel, a spring force and a rated life, and the three interact. A probe with a long travel tolerates more board variation and needs more room, and a probe with a high force makes a better contact and wears the pad faster.

Wear is counted in cycles rather than in hours, and the probes are replaced on a schedule. Where the pad is small, the probe tip is chosen to mark the pad rather than to cut into it, because a probe that cuts through the finish leaves a site that will corrode. Our flying probe notes describe the same contact problem without a fixture.

Access and the Keep-Out Area

The fixture needs access to the test points from a direction that the components do not block. That means the test points are placed on the side the fixture approaches, with a clear area around each one so that the probe can land without touching a neighbouring part.

Single sided access is the goal, because a fixture that loads the board from both sides costs twice as much and takes twice as long to load. Achieving it is a layout decision that has to be made early, before the components are placed.

Board Location and Repeatability

The board is located in the fixture by tooling holes or by its outline, and the accuracy of that location sets the accuracy of the probe contact. A board that is positioned by the edge only will move differently in each fixture, so the probes have to tolerate more error.

Tooling holes are the more repeatable option where they exist. Where they do not, the fixture uses a nest that contacts the outline at three points and clamps at one, and the clamping force has to be low enough not to bend the board.

Electrical Considerations

Long wires add inductance and capacitance, which matters for a high speed test and not at all for a continuity check. The fixture wiring is therefore matched to the test, and a fixture built for one test may not be usable for another.

Grounding is the second consideration. A common ground that is shared across many probes introduces coupling between signals, and it shows up as a test that passes on a bench and fails in production.

Maintenance and Changeover

The fixture wears in the probes, the wires and the mechanism, and the maintenance that keeps it working is scheduled rather than reactive. Recording the cycle count at each maintenance gives a trend that predicts the probe replacement rather than discovering it.

Where several products share a fixture, the changeover is a set of probe modules and a program rather than a new fixture. That arrangement is cheaper and it depends on the products sharing a test point pattern, which is a design decision. Our first pass yield notes describe how fixture related failures appear in the data.

Verification of the Fixture Itself

The fixture is verified with a known good board and a known bad board, so that both the pass path and the fail path are confirmed. A fixture that has only been proved with good boards is only half verified.

The known bad board is worth keeping for the life of the fixture, because it also detects a probe that has stopped making contact. Our board quality notes describe how the reference parts are controlled.

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. Where a value sits close to a process limit, the drawing should say so, since the shop can then open the process window rather than working to a nominal figure that carries no tolerance.

A stack-up that is drawn rather than described removes most of the ambiguity from a quotation, and it lets the fabricator price the board against the dielectric and copper weights that will actually be used. Where the requirement is not written down, the shop supplies its own default, and the default is chosen for the process rather than for the design.

Checks Before Release

The checks that matter are the ones performed on the product rather than on a sample kept for the purpose, because a coupon that travels with the panel is the only evidence about that panel. Where a requirement can be measured, it should be measured at the point of manufacture and recorded against the board or the lot it applies to.

A parameter that is set once and never re verified drifts, and the drift is usually discovered by a defect rather than by the record.

FAQ

Can a fixture test high speed signals? It can, and the wiring and grounding have to be designed for the signal rather than reused from a continuity fixture.

Does the probe damage the pad? It leaves a mark, and a probe that cuts through the finish leaves a site that can corrode. Tip geometry and force are chosen to leave a witness rather than a wound.

What does gopcb provide for test fixtures? We provide test point layout that gives single sided access, probe selection matched to the pad and the cycle count, fixture location designed for repeatability, wiring and grounding matched to the test, known good and known bad verification, and maintenance records by cycle count.