PCB Test Points: Design, Placement and Best Practices

PCB test points are the least glamorous features on a board and the ones most often deleted when space gets tight. That deletion is usually regretted, because a net that cannot be probed has to be diagnosed through the components attached to it, and finding a fault that way takes far longer than the board area the test point would have consumed.

What Test Points Are For

A test point is a defined landing area for an instrument probe or for a fixture pin. It exists so that a net can be measured directly, which means a fault can be isolated to a component rather than inferred from a system level symptom. It also gives the production line a repeatable way to verify every board without relying on operator judgement.

They are used at bring-up, where an engineer needs to see a supply rail or a clock, and at production test, where a fixture contacts dozens or hundreds of points at once, in a single operation that takes seconds per board. The difference in speed between a board that can be probed and one that cannot is measured in minutes per unit. The same feature serves both, but the physical requirements differ: an engineer with a handheld probe tolerates a small pad, while a fixture pin needs a defined geometry and clearance.

Test Point Types and Sizes

The simplest test point is a small pad on a net, either a dedicated pad or an existing via with an accessibly sized annular ring. A dedicated round pad of about 1.27 mm diameter gives a fixture pin a reliable target and is the standard for bed of nails probing. Smaller pads of 0.8 mm work for hand probing and for flying probe systems.

Vias can serve as test points if they are not tented by solder mask and if the annular ring is large enough for the pin to contact reliably. Component pads should generally not be used, because the probe can damage the joint or shift the part, particularly on small passives. Where a component pad is the only accessible point, it should be flagged as such and the fact noted in the assembly documentation.

<img src="https://www.gopcba.com/wp-content/uploads/2026/08/AVs-PCB.webp" alt="Test point pads arranged for probe access on a PCB” />

Placement Rules for Probe Access

Every test point needs clear space around it. A pad of 1.27 mm should have at least 2.54 mm between its edge and any adjacent component, so that the probe body and its guide can occupy the gap without touching a part or shorting a neighbouring net. Test points must also stay at least 3.2 mm clear of the board edge for the fixture plate.

Test points should be placed on the solder side wherever possible, since the component side is occupied. Where a net is only reachable on the top surface, a via placed beside it and left untented provides an accessible point on the underside. Planning this during placement costs nothing and reworking it later costs a new layout, as covered in escape routing and fanout rules.

Choosing Which Nets Deserve a Point

Not every net needs testing. The priorities are the supply rails, the ground reference, the reset and clock lines, the main data buses, and any net that carries a signal a functional test depends on. Nets buried under a package are usually the ones that most need a via-based test point, because they are hardest to reach any other way.

The count matters for cost. A fixture is priced by pin count and by density, so adding a hundred unnecessary points raises the non-recurring engineering charge directly. Test coverage should be decided deliberately, net by net, rather than by adding points to everything the tool allows. A short written list of the nets that must be probed, agreed before layout, prevents both over-provision and a late scramble to find space for a point that was forgotten.

<img src="https://www.gopcba.com/wp-content/uploads/2026/08/Turnkey-PCB-Assembly.jpg-1.webp" alt="Bed of nails fixture probing test points on an assembled board” />

Test Points and Signal Integrity

A test point is a small stub of copper attached to a net, and a stub is a discontinuity. On a low speed signal the effect is negligible, but on a high speed differential pair or a fast clock, the added capacitance loads the line and the stub can create a resonance that degrades the eye diagram. Where the signal must be probed, keep the stub short or place the pad directly on the trace rather than at the end of a branch.

Where the frequency is high enough that any stub is unacceptable, the test point should be omitted and the net verified by other means, such as a functional test or an on-board monitoring circuit. A useful compromise is to place the pad so that it can be depopulated in the final artwork once the design is proven. For differential pairs, a symmetric pair of pads can be used as a differential landing area, which produces far less disturbance than two single ended probes.

Test Points on Assembled Boards

Once components are fitted, the test point population is what a bed of nails fixture contacts, and it is the only interface the fixture has to the assembly. The fixture pins have a limited travel and a defined force, so the pads must be flat, free of solder and free of residue. Solder mask must not encroach on the pad opening, and no component or tall hardware may sit inside the pin field.

Test points also matter for rework. A board that has been repaired has to be retested, and the retest uses the same fixture. If a rework operation covers a test point with adhesive or coating, the board can no longer be verified, which is a common problem on assemblies that are coated after test, described in board outline and mounting design.

Documenting and Numbering Test Points

Each test point should appear in the assembly documentation with a reference designator, its net name and the layer it is on. A numbered sequence helps the operator follow a test procedure, and it lets the fixture programme be checked against the layout. Without that documentation, a test point is just an anonymous pad.

Marking the pads as a distinct object type in the CAD database makes them auditable. They can be counted, checked for clearance automatically and moved as a group when the layout changes. That discipline is part of the wider set of design quality practices described in PCB design quality characteristics.

FAQ

How large should a test point be? For fixture probing, 1.27 mm diameter is a safe standard with 2.54 mm of clearance, and it can be reduced to 0.8 mm for flying probe or manual probing. Below that, contact becomes unreliable across a production run.

Can a mounting hole be used as a test point? Only if it is connected to the net being measured, which is unusual. Ground is the exception, since chassis and mounting points are often tied to ground, and a ground reference nearby is genuinely useful for a probe.

Do test points need to be removed before shipping? No. They are copper features under the solder mask or exposed as defined pads, and they remain part of the finished board. Nothing needs to be changed between the tested unit and the shipped unit.