How Engineers Should Evaluate Spurious Performance in RF Signal Generators

One spurious number is not enough. Harmonics, sub-harmonics, non-harmonic spurs, output level, and real RF test workflow all matter.

How Engineers Should Evaluate Spurious Performance in RF Signal Generators

Spurious performance is one of those RF signal generator specifications that looks simple at first.

It is often shown as one number.

But in real RF and microwave test setups, that number usually needs more context.

A spurious signal may appear as a harmonic, a sub-harmonic, or a non-harmonic spur. Each type can affect a measurement in a different way. For receiver testing, mixer testing, RF component characterization, radar-related evaluation, and automated RF test workflows, the question is not only whether the source looks clean in the datasheet.

The better question is:

Can the unwanted signal affect the actual measurement?

Spurious performance is different from phase noise

Phase noise and spurious performance are both related to signal purity, but they describe different issues.

Phase noise is usually evaluated close to the carrier and read across offset frequencies.

Spurious signals are discrete unwanted signals that appear at specific frequency points.

In a spectrum view, phase noise looks like energy spreading around the carrier. Spurs often look like narrow lines.

This matters because the measurement risk is different. A receiver may be affected by a narrow spur if it appears inside a sensitive passband, even if the general noise behavior looks acceptable.

Not all spurs are the same

Engineers should separate harmonics, sub-harmonics, and non-harmonic spurious signals.

Harmonics are unwanted signals at integer multiples of the carrier frequency. They can matter when testing amplifiers, filters, mixers, and frequency converters.

Sub-harmonics may appear because of frequency division, multiplication, or conversion behavior. They are worth checking in wideband microwave signal generation.

Non-harmonic spurious signals are not simple multiples of the carrier. They may appear at unexpected positions or offsets. These are often important in receiver-related testing because they may fall into a sensitive measurement band.

This is why I do not think one “spurious” number is enough for serious RF source selection.

The test condition matters

Spurious performance depends on operating condition.

Carrier frequency matters.

Output level matters.

Operating mode matters.

A value measured at 10 GHz does not automatically describe behavior at 20 GHz or 40 GHz. A value measured at one output level may not describe performance at another output level.

Engineers should check the condition behind the number: carrier frequency, output level, CW or modulation mode, and whether the value refers to harmonics, sub-harmonics, or non-harmonic spurs.

Without that context, the specification is hard to apply to a real test plan.

Think from the DUT plane

The most useful way to evaluate spurious performance is to start from the device under test.

For receiver sensitivity testing, a small unwanted signal can create confusion if it appears near the receiver passband.

For mixer testing, source spurs can interact with LO drive, RF input, IF output, and conversion products.

For amplifier or RF module testing, harmonic content may affect output spectrum analysis.

So the useful question is not only:

“What is the spurious level?”

A better question is:

“Can this unwanted signal affect my measurement at the DUT plane?”

Sweep, pulse, and modulation workflows need more attention

Spurious evaluation is not only a CW problem.

In sweep applications, the carrier moves across a frequency range. Harmonics and other unwanted products may move as well. If the DUT is sensitive in part of that range, the test result may be affected.

In pulse-related testing, timing behavior, on/off behavior, and spectral content should be considered together.

In analog modulation workflows, unwanted signals should be evaluated together with carrier frequency, output level, modulation behavior, and receiver response.

This is why RF signal generator selection should be based on the full test workflow, not only the maximum frequency.

Select by workflow fit

Spurious performance is important, but it should not be evaluated alone.

A practical RF signal source should match the frequency range, output power, phase noise, spurious performance, level accuracy, sweep behavior, pulse or modulation needs, reference requirements, and remote-control workflow of the actual application.

For automated RF test systems, repeatability may matter more than one attractive number.

For receiver testing, unwanted signal control may be critical.

For RF component characterization, clean and predictable output makes the result easier to trust.

Before comparing RF signal generators, I would ask:

Does the datasheet separate harmonics, sub-harmonics, and non-harmonic spurious signals?

Are the test conditions clearly stated?

Is the carrier frequency relevant to the planned test?

Is the output level close to the real operating condition?

Could any unwanted signal fall inside the DUT’s sensitive band?

Does the workflow use CW, sweep, pulse, or modulation?

Will the source be used manually or integrated into an automated test system?

These questions usually provide a better selection path than comparing one headline number.

Original technical article published by Corech: https://www.corechmw.com/News/rf-signal-generator-spurious-performance/

Related product category: https://www.corechmw.com/rf-signal-generator/

Related signal generator platforms: https://www.corechmw.com/rf-signal-generator-sweep-pulse-40ghz-csg9k40ga/

https://www.corechmw.com/rf-analog-signal-generator-43ghz-cams9k40g/