Correcting the Measurement Path for RF Harmonic Measurements

A harmonic measurement spans different frequencies, so the measurement path can attenuate the fundamental and harmonics by different amounts. A single cable-loss correction can therefore produce an incorrect harmonic ratio. At the same time, a strong fundamental can cause distortion inside the analyzer. Review both effects before accepting the result.

Map the frequency-dependent path

The Rohde and Schwarz note on harmonic measurements explains analyzer dynamic-range considerations and the use of high-pass filtering to reduce the fundamental during harmonic measurements.

List cables, pads, filters, couplers, and adapters between the device and instrument. Characterize their transmission at the fundamental and at every harmonic being reported.

Apply corrections to each frequency

Convert measured levels to the same device reference plane using the relevant loss for each frequency. Then calculate harmonic level relative to the corrected fundamental if the result is reported in dBc.

For example, a path with 1 dB loss at the fundamental and 3 dB at the second harmonic makes the raw second-harmonic ratio appear 2 dB better than the ratio at the device output. A common 1 dB correction does not remove that error.

Protect the analyzer’s dynamic range

Check the fundamental level reaching the analyzer input and its internal measurement chain. Use appropriate attenuation or filtering while keeping the harmonic above the measurement floor.

If a filter is added, measure its loss and verify that its input match and power handling are suitable. A filter that changes the load seen by the device can alter the harmonic generation being measured.

Test whether the result belongs to the device

Change analyzer input attenuation or a characterized external pad while preserving the device operating point. After correcting for loss, a genuine harmonic should remain consistent within the measurement uncertainty.

Measure the source spectrum when testing an amplifier. Harmonics already present at the input can contribute to the output and should not automatically be attributed to amplifier nonlinearity.

Record detector, resolution bandwidth, averaging, and frequency span. Confirm that the settings resolve the intended harmonic and do not mix it with nearby unrelated spurs.

Repeat at the required output powers, frequencies, and loads. Harmonic performance can vary across those conditions even when the fundamental gain is nearly unchanged.

Save per-frequency path corrections, analyzer-floor checks, source evidence, and corrected levels. The report should make its reference plane and dBc denominator explicit. A small marker reading is useful only after the measurement path and the analyzer’s own distortion have been accounted for.