Impedance Coupon: Preparation, Placement and Process Control

A controlled impedance board is only as good as the evidence that it meets the requirement. That evidence is normally a coupon built on the same panel with the same stackup, carrying traces whose geometry is known, so that a measurement can be compared against the target value.

An impedance coupon is therefore a design task in its own right. The trace geometry has to match the production traces, the reference plane has to be the same, and the coupon has to be placed where it will be processed identically to the product it represents.

Why Coupons Are Built

Impedance depends on the trace width, the dielectric thickness, the dielectric constant and the geometry of the return path. All four vary slightly in production, so the impedance of a real trace can only be confirmed by measuring something that was built under the same conditions.

The coupon makes that possible without destroying product. It is usually placed in the panel border or in a breakaway area, and it carries the same layer structure as the boards around it. After measurement the coupon is discarded together with the border. The stackup the coupon reproduces is described under layer stackup from one to eight layers.

Test Trace Geometry

A test trace is designed to be measurable, not to be representative of the most difficult net. It is usually a straight run of several centimetres, long enough for a time domain reflectometer to resolve the impedance and short enough to fit inside a coupon. Bends and layer changes are avoided because they add reflections.

Where several impedances are required, one test trace is provided for each. The traces should be spaced far enough apart that they do not couple to each other, and any ground pour adjacent to them should match the pour on the production board.

Reference Planes And Grounding

The reference plane under the test trace must be continuous and must be connected to the measurement ground with a low impedance path. A plane that is split under the trace changes the return path and therefore the measured impedance, and a plane that is poorly grounded introduces an inductance that corrupts the result.

Grounding is usually provided by vias along the side of the coupon and by a wide ground pad at the probe point. The via spacing should be small compared with the wavelength of the highest frequency in the measurement, which for a typical time domain reflectometer means a few millimetres.

Coupon Placement On The Panel

The coupon should sit in the same area as the product, not at a corner on its own, because dielectric thickness varies across a panel. Placing it next to a dense area reproduces the local conditions, while placing it in an empty region can give a result that does not represent the boards.

Some designs place a coupon at both ends of the panel to show the spread across the panel. That costs a little more border area but gives the fabricator information that is useful when a measurement comes back marginal.

Measurement And Acceptance

Measurement is normally made with a time domain reflectometer, which sends a fast edge along the trace and reports the impedance as a function of distance. The trace is probed at a defined point with a controlled launch, and the result is averaged over a region that excludes the ends.

Acceptance limits are usually plus or minus 10 percent of the target, and the coupon result is recorded with the lot. Where a measurement falls outside the limit, the fabricator can often adjust the trace width in the artwork to bring the impedance back to target. The way these requirements reach the fabricator is described under PCB design and fabrication.

Common Mistakes

The most common mistake is a coupon that does not match the product. A test trace on a different copper weight, a different dielectric thickness or with a split plane produces a number that has nothing to do with the board being shipped, and it can send both parties looking for a problem that does not exist.

The second is measuring at the wrong point. Probing near a pad, a via or a connector introduces a discontinuity that is read as an impedance change, and the resulting curve is then interpreted as a process problem. The launch and the measurement window should be defined on the coupon drawing. How the coupon is checked as part of the build record is described under PCB design quality characteristics.

Process Control and Verification

Running a first article through the same checks as the production panel confirms that the two agree, and that comparison is the cheapest form of process control available at prototype stage. Keeping a sample from the panel turns a dispute into a measurement, because the same coupon can be re-examined by both parties without rebuilding the batch.

Reviewing the design before the data is released is cheaper than correcting it after the panel is in the tank, because every step downstream inherits the decision made at the front end.

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

Does every controlled impedance board need a coupon? In practice yes, because the impedance cannot be verified from the completed product without destructive measurement. The coupon is the practical way to provide evidence without scrapping boards.

How long should a test trace be? Long enough to resolve the impedance, typically 50 to 100 mm, and short enough to fit the coupon. Longer traces make the launch effects less significant inside the measured window.

Can impedance be calculated instead of measured? It can be predicted from the stackup, and that prediction is used to set the trace widths. Measurement is still needed to confirm what the process actually produced.