GHz

In the era of 5G, Wi‑Fi 6/7, and next‑generation high‑speed interfaces such as PCIe 5.0 and USB4, the operating frequencies of consumer and industrial electronics have migrated squarely into the gigahertz band. At these frequencies, the physics of electromagnetic propagation changes dramatically, and the conventional low‑frequency EMC test methods no longer suffice. This article explores the unique challenges of radiated emission testing in the GHz regime and outlines practical engineering strategies to overcome them.

The relationship between wavelength (λ) and frequency (f) is inversely proportional:

[
\lambda = \frac{c}{f}
]

where c is the speed of light. When f climbs into the gigahertz range, λ shrinks to centimeters or even millimeters. This compression has two profound consequences:

Device dimensions become comparable to the wavelength. Small gaps, seams, and PCB traces can act as resonant antennas.
Signal attenuation and dispersion increase. Transmission lines and cables exhibit higher loss, and reflections become more pronounced.

Because of these effects, the radiated emission measured during EMC testing may no longer reflect the true behavior of the equipment under test (EUT). Instead, the test environment and measurement chain can introduce significant artifacts.

Transmission Line and Cable Issues

Cable as an Unwanted Antenna

At low frequencies (30 MHz – 1 GHz), cables are often treated as passive conduits. In the GHz domain, however, the coaxial shield can become an efficient radiator if the common‑mode current is not properly suppressed. Poor grounding of the shield or inadequate shielding effectiveness can cause the cable itself to dominate the emission spectrum, leading to a false fail.

Rapidly Rising Insertion Loss

Insertion loss (IL) grows steeply with frequency. A standard RG‑58 coax might exhibit only a few dB of loss at 1 GHz, but it can exceed 10 dB per meter above 3 GHz. Without a precise IL correction, the measured field strength will be underestimated, masking real compliance issues.

Standing Waves and Reflections

Impedance mismatches between the cable, connectors, and the test receiver produce standing waves. These waves manifest as periodic peaks and valleys in the spectrum, obscuring the true emission profile. The situation is exacerbated when the cable length is not an integer multiple of half‑wavelengths.

Practical Tips

Use low‑loss, double‑shielded cables (e.g., LMR‑400) with verified attenuation curves up to the target frequency.
Perform a full‑band IL calibration for each cable run, applying the correction factor to the measured data.
Verify impedance matching at both ends using a time‑domain reflectometer (TDR) or vector network analyzer (VNA).

EUT Structural Leakage and Coupling Paths

Slot and Gap Radiation

When a physical gap approaches λ/2 or λ/4, it behaves as a slot antenna. For example, at 10 GHz (λ ≈ 3 cm), a 1.5 cm seam can radiate strongly. Even a loose screw or a small crack in the chassis can become a significant source of interference.

Connector Shielding Failures

High‑frequency signals are highly susceptible to mode conversion at connector interfaces. If the connector shield does not achieve a 360° low‑impedance bond with the chassis, the interface can leak substantial energy. This is often overlooked during design but becomes critical in the GHz band.

PCB Trace Emission

Differential pairs that are not perfectly matched or that lack a continuous return path generate common‑mode currents. Microstrip lines, strip‑line, and via‑holes can act as miniature antennas. At gigahertz frequencies, even a few millimeters of asymmetry can produce measurable radiation.

Illustrative Case

A high‑speed switch passed all tests below 1 GHz but exceeded limits at 5 GHz. Investigation revealed that the heat‑sink perforations formed a periodic array, creating a constructive interference pattern at that frequency. Redesigning the perforation pattern eliminated the issue.

Test Environment and Antenna Measurement Challenges

Absorber Performance

Antenna absorbers in anechoic chambers are frequency‑dependent. At very high frequencies, the material may not absorb sufficiently, allowing reflections from walls and floor to interfere with the direct signal. This can produce large measurement swings.

Antenna Polarization and Directivity

The polarization of the test antenna must match that of the EUT’s dominant radiation mode. Mis‑alignment can lead to a false pass because the antenna is insensitive to the emitted field. Moreover, many commercial antennas exhibit significant beam distortion above 3 GHz, altering the measured gain.

Near‑Field vs Far‑Field

The far‑field distance ( R = \frac{2D^2}{\lambda} ) (where D is the largest dimension of the antenna) increases as λ decreases. In a typical anechoic chamber, the far‑field zone may be unreachable for frequencies above 10 GHz, forcing the use of near‑field techniques or extrapolation methods.

Receiver Dynamics and Noise Floor

Elevated Noise Floor

Spectrum analyzers and receivers designed for lower frequencies often have higher noise floors in the GHz range. If the EUT’s emission is close to the noise floor, the measurement may be unreliable. Using a receiver with a lower noise figure or extending the averaging time can mitigate this.

Image and Spurious Signals

Mixers inside the receiver can generate image frequencies and spurious tones. At gigahertz frequencies, these artifacts are more likely to fall within the measurement band, potentially masking real emissions or creating false positives.

Comprehensive Mitigation Strategy

Focus Area	Key Actions
Shielding	• Apply conductive gaskets to all seams. • Ensure 360° grounding of connector shields. • Use double‑shielded cables with verified attenuation.
PCB Design	• Maintain differential pair symmetry. • Provide a continuous ground plane. • Add common‑mode chokes at high‑speed interfaces.
Calibration	• Build a full‑band cable loss model. • Calibrate antenna factor across the target band. • Perform periodic TDR checks for impedance integrity.
Test Environment	• Verify absorber effectiveness at the highest test frequency. • Use a calibrated, well‑characterized test antenna. • Apply near‑field to far‑field transformation when far‑field is unattainable.
Measurement Practices	• Extend averaging times to lower the effective noise floor. • Use image‑rejection techniques or dual‑mixing schemes. • Cross‑check results with a second receiver or a calibrated reference source.

By systematically addressing these areas, engineers can significantly reduce the uncertainty inherent in GHz‑band radiated emission testing. The result is a more reliable assessment of compliance, shorter EMC debug cycles, and a smoother path to market for high‑frequency products.

Takeaway

The gigahertz band introduces a host of new variables that can skew EMC measurements. Recognizing the role of cable radiation, structural leakage, environmental reflections, and receiver limitations is the first step. From there, a disciplined approach—combining robust shielding, meticulous PCB layout, rigorous calibration, and a verified test environment—provides the foundation for accurate, repeatable radiated emission testing in the GHz domain.