PCB Impedance Test Coupons and TDR Testing

A test coupon is a small structure, usually a straight trace segment with a defined launch point, built with the same copper weight, dielectric, and reference-plane construction as one impedance group on the production board. It is not a cut-out of the actual board; it is a separate structure panelized in the scrap area around the production boards so it can be probed or clipped for measurement without damaging a shippable unit.

Each distinct impedance group—different target, different layer, different structure—generally needs its own coupon, because the whole point is measuring that specific geometry and stackup combination, not a generic board property. A panel with single-ended 50 Ω microstrip on L1 and 100 Ω differential stripline on L3 needs two coupon designs, not one.

Dielectric thickness and etch factor are not perfectly uniform across a lamination panel—there is normal process variation from center to edge and between panels in a lot. A coupon placed at the panel edge measures edge conditions; a coupon placed near the board cluster measures conditions closer to where the actual boards sit. Neither placement is wrong, but the report should state where the coupon sat so a reviewer can judge how representative the number is.

For tighter tolerances (±7% and below), it is common to request multiple coupons per panel, or a coupon placement agreement that puts coupons adjacent to the specific boards being measured, rather than one coupon per panel regardless of where the boards sit.

Time-domain reflectometry sends a fast voltage edge into the coupon trace through a controlled launch (typically an SMA connector or probe point) and measures the reflections that come back as the edge encounters impedance discontinuities along the line. Because the measurement is time-resolved, it produces an impedance profile along the length of the trace, not a single averaged number—which is useful for spotting a localized defect, like a via stub or a connector launch mismatch, that a single-point measurement would hide.

The reported value is usually read at a stated point along the flat section of the trace, away from the launch discontinuity and the open or terminated end, where the measurement best represents the steady-state trace impedance. A report that quotes “the TDR number” without saying where on the trace it was read is missing a detail that matters for comparing results across coupons.

At minimum: the coupon or group ID matching the impedance table, the target value and tolerance, the measured value, pass/fail against tolerance, the launch method, the equipment used, and the measurement date. For tighter-tolerance or RF-adjacent work, the full impedance-vs-distance trace (not just the single reported number) lets a signal-integrity engineer verify there is no localized anomaly even if the summary number passes.

If your program needs retained records for supplier audits or field-failure investigations, specify that in the RFQ—retention and report format are agreement items, not defaults, and asking after the fact usually means the records were not kept in the format you need.