Flying Probe Testing vs Fixture Test: 6 Rules to Choose

Flying probe testing and fixture testing both check that a bare board matches its netlist, and they get there in completely different ways. A prober moves a small number of heads across the panel and measures nets one at a time, while a fixture presses hundreds of pins against a dedicated bed of nails and tests many nets at once.

The choice is usually presented as a question of volume, but the real answer depends on three things: how many boards will be tested, how much test access the design provides and what the test has to measure. Getting the sequence wrong costs either money in fixtures or time in test.

What Electrical Test Has to Prove

Bare board test opens and shorts. It confirms that every net is continuous and that no net is connected to another, including the ones that are only reachable through a via or on an inner layer. Everything else, such as impedance and resistance measurements, is a bonus the same equipment may or may not deliver.

Because most of a board is hidden inside the stack, test is the only operation that examines inner layer integrity across the whole panel. That is why the netlist has to be verified against the original design data, so the test is checking the board rather than the exported file.

How Flying Probe Testing Works

Moving probes are programmed from the CAD data and address a defined set of test points on both sides of the board. Because the heads move, the number of physical contact points is small, and no fixture has to be built, so the setup cost is low and the program can be changed by editing data.

The cost is time. Each measurement takes a contact, a settle period and a reading, so test time grows with the number of nets. That makes flying probe attractive for prototypes and small batches, and increasingly expensive as volume rises.

How a Fixture Works

A fixture is a plate carrying spring probes positioned to match the test points on the board. All the probes contact at once, so hundreds or thousands of nets can be measured in a fraction of the time a prober needs. The fixture is built specifically for the board and is only useful for that design.

Its costs are front loaded. The fixture has to be designed, drilled and assembled, and the lead time for a new fixture is measured in days. Once it exists, the cost per board falls quickly, which is what makes fixtures the standard answer for volume production.

Test Access and Net Coverage

Both methods need somewhere to make contact. A test point on a pad, a via or a connector pin is a place a probe can reach, and the test point design decides how much of the board can be tested by either method. Nets with no accessible point cannot be verified directly.

Fixtures are more sensitive to access geometry, because the probes are fixed in position and have to sit flat on their target. A prober can sometimes reach a point at an angle or land on a small feature, which gives it an access advantage on dense layouts. Neither can test a net that has no landing site at all.

Volume, Setup and Cost per Board

The crossover is arithmetic. Fixture cost divided by batch size gives a per board figure that falls steeply, while flying probe time multiplied by the hourly rate gives a per board figure that stays flat or rises with net count. Estimating both before committing is a five minute exercise that prevents an expensive fixture on a short run.

Repeat orders change the arithmetic. A design that will run for years justifies a fixture even at moderate volume, while a design that will be revised in a month does not. Where a product family shares a panel, the fixture can often be reused across variants, which moves the crossover point again.

Limits: Resistance, Capacitance and Components

Both methods measure resistance, and both can be fooled by a resistive short through a partly etched sliver or a plating defect. Four-wire measurement, using separate drive and sense connections, is needed for low resistance values, and its availability depends on the machine and the test setup.

Capacitance testing looks for opens on nets that have no accessible point at both ends, by measuring the coupling between a net and a reference plane. It is slower and needs a stable setup, so it is usually reserved for the nets that cannot be reached any other way.

When Both Are Used Together

The two methods are often used in sequence. A flying prober validates the first articles and the pilot batch while the fixture is being built and verified, then the fixture takes over for production. The prober then stays useful for engineering changes and for boards returned from the field.

They are also complementary in coverage. Where a fixture cannot reach a particular net, the prober can often measure it in a separate pass, so the total coverage is higher than either method achieves alone. The in-circuit test stage on an assembled board may then inherit a known-good set of nets from both.

Test Point Design and What the Fab Needs

Test points should be specified on the fabrication drawing with a size, a clearance and a grid that the equipment can use. A test point that is too small or too close to a neighbouring pad forces the prober to slow down or the fixture to use a finer probe, and both raise cost.

Providing the netlist in the correct format and identifying which nets are critical for capacitance testing is part of the data package. Where the design leaves nets inaccessible, the drawing should say so, because a shop that discovers it at programming time will either quote a premium or leave those nets untested. The engineering test station is where this review usually happens.

Choosing for New Products and Production

New products belong on a prober until the design is stable. Engineering changes, revisions and short builds all favour a method that needs no tooling and can be reprogrammed from data, and the additional test time is a small price for that flexibility.

Once the design settles and volume grows, a fixture becomes the economical choice, and the decision should be revisited whenever the volume forecast changes. A fleet that is all probers is slow and expensive at volume, while a fleet that is all fixtures cannot support the prototypes that feed the production line.

FAQ

Is flying probe testing less accurate than a fixture? The measurement principle is the same, so accuracy is comparable when both are set up correctly. The practical difference is speed and repeatability of contact, and a fixture with worn probes can be less reliable than a prober that verifies contact at every point.

Can a fixture test a board that has no test points? Not usefully. A fixture needs physical access to every net it measures, so a design with no test points can only be tested in the areas that remain accessible, and the remaining nets have to be covered by capacitance methods or accepted as untested.

When does a fixture become worthwhile? When the expected number of boards is large enough that the fixture cost divided by the volume is less than the added test time on a prober. That crossover depends on net count and test time, and it should be calculated rather than assumed.