Make Near-Field EMI Probe Comparisons Repeatable

A smaller peak after a board modification is useful only if the measurement geometry remained comparable. Near-field probes respond strongly to where and how they are held. Establish a repeatable comparison before using a handheld scan to decide which layout or component change worked.

Choose the local question

Tektronix’s EMI troubleshooting note describes near-field probes as tools for relative measurements and emphasizes distance and orientation. It distinguishes electric-field and magnetic-field probes. Their readings help locate activity; they do not directly replace a calibrated radiated-emissions measurement.

State the frequency or event being investigated and the board operating mode that produces it. Mark candidate source regions on a photograph. Keep the initial broad scan separate from the more controlled before-and-after comparison used to evaluate a change.

Fix the geometry

Use a nonconductive positioning aid or another reproducible method to locate the probe. Record height, orientation, probe model and the exact point on the board. A photograph from two directions can be more informative than a note saying above the regulator.

Keep probe cables and the product’s external cables in the same arrangement. Changing a cable route while replacing a component introduces another variable. If a larger probe identifies a region and a smaller probe refines it, establish a new baseline with the smaller probe before comparing amplitudes.

Preserve the operating and instrument state

Record input supply, load, firmware mode, clock settings and any activity pattern used during the scan. Retain analyzer bandwidth, detector, acquisition time, attenuation and preamplifier state. A peak captured during a short burst may be missed if the acquisition does not cover the same event.

Before modifying the board, remove and replace the probe several times and observe the repeatability. This gives a practical indication of how much variation the positioning method introduces. Do not present a small change as decisive when it is comparable to that observed variation.

Confirm the effect beyond the local scan

After a candidate change reduces the local reading, check the relevant cable or system-level measurement. A source can couple through more than one path, and reducing one local field does not guarantee the same improvement at the product’s external interface.

Retain the baseline, modified and restored-baseline results when a reversible experiment allows it. Explain which frequency and condition improved and which measurements remain unchanged or untested. This turns the probe into a controlled diagnostic tool and prevents a different hand position from being mistaken for evidence that a PCB redesign solved the emissions problem.