Preparing a Two Port Shunt Measurement of PCB PDN Impedance

Power distribution networks can have impedances far below a network analyzer’s reference impedance. A two port shunt arrangement can be useful for this measurement, but the fixture and return connections become part of the experiment. Establish the electrical model before converting a transmission trace into a PDN impedance plot.

Choose the appropriate method

The Rohde & Schwarz application note on PDN impedance compares reflection, series, and shunt measurement arrangements. The two port shunt method derives impedance from transmission through a connection in which the low impedance DUT shunts the signal path.

Use the equation and assumptions appropriate to the actual fixture and reference impedance. A conversion intended for an ideal symmetrical connection should not be applied blindly to arbitrary cable and probe arrangements. Document any fixture correction and the frequency range over which it has been checked.

Define exactly what is connected

Identify the rail node, return node, attachment locations, and current path through the fixture. Photograph both signal and return connections. Two probes touching the same nominal power net at distant locations may include different spreading paths from a compact connection at one location.

List which capacitors, regulators, connectors, and loads remain attached. An unpowered assembled board, a bare capacitor network, and an operating rail are different DUT states. Do not use one result as an automatic description of the others.

Protect the instrument and validate the fixture

Review port voltage and power limits before connecting to an energized board. Use a measurement arrangement specifically suitable for any DC bias involved, including appropriate blocking or bias handling where required. Never assume that an RF port can tolerate the rail simply because the expected AC signal is small.

Check the fixture with suitable known impedances or reference structures over the intended range. Repeat connections and compare results. Inspect whether cable positioning or return routing changes the apparent resonance; strong sensitivity indicates that the setup requires further characterization.

Interpret the curve with its boundaries

Save raw transmission data as well as the converted impedance. Record source level, frequency grid, IF bandwidth, calibration, board state, and conversion method. Compare repeated measurements before assigning significance to a small peak or dip.

A resonance in the measured network is a starting point for investigation, not a direct prediction of rail excursion under every load step. Relate the impedance result to the actual load spectrum, operating condition, and observation point. This connection makes the measurement useful for selecting a concrete decoupling or layout experiment.