Near Field, Far Field and RFI: Stop Using Wavelength as a Verdict
Near Field, Far Field and RFI: Stop Using Wavelength as a Verdict
A 160 m station can put much of a house close to an antenna in wavelength terms. That does not make every low-band EMC problem “near field,” or every 2 m problem “far field.” Find the source, geometry and current path first.
“How many wavelengths away?” is useful only after “which electromagnetic source?” The transmitting antenna has field regions. So can a feed line carrying exterior current, a switching loop, a power lead, an enclosure seam or a long data cable. One room may contain several overlapping fields and a conducted path at the same time.
This is why the easy band rule fails. Low frequency makes a given physical distance electrically shorter, but source size, source impedance, current distribution, cable routing, nearby structures and the victim circuit still decide how interference couples. The fix must follow that path—not a label attached to the band.
There Are More Than Two Useful Regions
Close to a radiating structure, stored electric and magnetic energy can dominate. The local E/H amplitude and phase relationship can change sharply with position, and neither field component can safely be inferred from the other. This is the reactive near field.
Farther away, radiation dominates but the angular field distribution may still change with distance. This radiating near field, often called the Fresnel region, is particularly important for electrically large antennas and apertures. Only in the far field is the angular pattern essentially independent of distance and the wave locally approaches plane-wave behaviour. In free space, E/H then approaches the intrinsic impedance of about 377 Ω.
The transitions are gradual, not walls painted in the air. Different standards use different conservative boundaries for their own measurement purpose. For example, ITU-T K.61:2025 uses field-region conventions for RF-exposure assessment, including antenna dimension D, wavelength λ and the familiar 2D²/λ aperture criterion. Those assessment boundaries are useful, but they are not a universal table that can be indexed by amateur band while ignoring the radiator.
| Region or path | What changes | What to measure |
|---|---|---|
| Reactive near field | Stored energy is important; local E/H ratio and phase depend strongly on source and position. | Measure electric and magnetic components separately when either may matter. |
| Radiating near field | Radiation dominates, but field structure and angular distribution can still vary with distance. | Use a method valid for the antenna geometry and measurement distance. |
| Far field | The local wave is approximately transverse and the angular pattern is stable with distance. | A calibrated field component can support plane-wave inference when the assumptions hold. |
| Conducted path | Disturbance voltage or current travels on power, signal, shield or bonding conductors. | Measure the relevant differential or common-mode voltage/current at declared reference points. |
A House Is Not One Field Region
On 160 m and 80 m, the wavelength is large enough that a house and garden can be electrically close to the station antenna. That makes reactive and induction coupling plausible, but it does not prove that they dominate. A coax exterior carrying common-mode current may be a second radiator. A mains lead may conduct the disturbance. A nonlinear receiver stage may create products that were never present in the air.
On 20 m, 10 m or 2 m, a nearby USB cable, router lead, microphone cable or switching-current loop can still be in its own near field. Higher frequency does not automatically turn the case into far-field radiation. Conversely, a distant HF transmitting antenna can illuminate the house as a far-field source while a local cable converts that field into a common-mode current near the victim.
Common mode and near field are also not synonyms. Common mode describes a voltage or current relative to a third reference or return structure. Near field describes spatial field behaviour around a source. A common-mode cable current can create a near field close to the cable and radiation farther away; it can also arrive through direct conduction.
Electric and Magnetic Clues Are Local, Not Identities
A compact high-voltage, high-impedance structure often produces an electric-field-dominant local region. A compact high-current, low-impedance loop often produces a magnetic-field-dominant local region. Those are useful diagnostic tendencies, not permanent identities of a device or antenna.
A small E-field probe can help locate voltage-rich structures. A small H-field loop can help locate current loops. Probe orientation, shielding, cable pickup, loading, bandwidth and calibration all matter. A strong indication at one point identifies a local coupling opportunity; it does not yet prove which path reaches the receiver.
IEC 61000-4-39 exists precisely because equipment exposed to an RF transmitter in close proximity needs an immunity method different from a simple uniform far-field assumption. That does not turn a hand-held near-field probe into a compliance instrument. Use probes to form and test a hypothesis, then verify the entire source–path–victim chain.
Why the Usual Fixes Sometimes Miss
- A choke addresses a current mode, not a field-region label. It helps only when it presents useful complex impedance in the actual common-mode path at the frequencies involved, without saturation, heating or a bypass path.
- A differential filter works between the intended conductors. It may reduce ripple while leaving a capacitively driven common-mode path almost unchanged.
- Shielding must interrupt the relevant coupling. Seams, apertures, connectors, cable terminations and penetrations can dominate an otherwise conductive enclosure.
- Distance and routing change mutual coupling. They can help in reactive, radiating-near and far-field cases, but the result depends on orientation and surrounding conductors.
- Bonding has both EMC and safety constraints. Never lift protective earth or improvise mains bonds to improve a trace. High-frequency bonding geometry and the legally required protective system are related engineering questions, not interchangeable ones.
A quarter-wave stub can be a powerful narrowband network element. Ferrite can add useful broadband loss or reactance to a selected mode. A bonded enclosure can control an aperture or cable reference. None works because the band has been declared “near field” or “far field.” It works because its impedance, placement and geometry change the measured path.
Diagnose the Path Before Buying the Cure
Record receiver, antenna, frequency, mode, bandwidth, gain/attenuation state, displayed level and the exact symptom.
Use safe power isolation, distance, orientation, E/H probes and clamp-current measurements to test one path at a time.
Return to the baseline after each change. A repeatable A/B/A result is stronger than a one-time disappearance.
- Rule out receiver-generated effects. Add attenuation, change preselection, compare another receiver or use a shielded termination. Overload and intermodulation can imitate an external emitter.
- Capture the signature. Record frequency spacing, occupied bandwidth, time behaviour and operating state. A spectrum without settings and reference level is a picture, not a measurement.
- Change distance and orientation. A rapid spatial change suggests local coupling, but reflections and cable rerouting can complicate a simple inverse-distance story.
- Compare E, H and cable current. The measurements answer different questions. Agreement between them can reveal the local source, exported current and receiving structure.
- Test the permitted path. Disconnect only safe low-voltage accessories, reroute one cable, or use normal circuit controls. Do not open hazardous equipment or defeat a protective conductor.
- Verify normal operation. Repeat the original receiver test over the affected frequencies and device states, then confirm safety, data integrity and thermal behaviour.
Field rule worth keeping: if the result depends strongly on exactly where a small E- or H-field probe sits, treat it as local evidence. If a clamp shows current on a cable, follow the cable. If a calibrated antenna sees a stable angular field at a valid distance, use the appropriate radiated-field method. Never substitute one observation for another.
Primary Standards and Engineering Guidance
- ITU-T K.61:2025: current field-region and measurement guidance, including reactive, radiating-near and far-field distinctions.
- ITU-T K.91:2024: current definitions and assessment guidance for non-uniform fields and separate E/H evaluation in the reactive near field.
- ITU-T K.37:2024: EMC installation practice for cabling, bonding, screening, separation and common-mode coupling.
- IEC 61000-4-39:2017: immunity testing for radiated RF fields from transmitters used in close proximity.
- NTIA TM-13-489: antenna models and near-/far-field parameters used in electromagnetic-compatibility analysis.
- NIST, Near-Field Scanning Measurements: rigorous near-field measurement concepts and transformation to antenna radiation information.
Mini-FAQ
- Is everything within one wavelength in the near field? No. Field regions depend on source geometry, electrical size and the purpose of the boundary. One wavelength is a conservative convention in some assessment methods, not a universal physical wall.
- Does near field mean there is no radiation? No. The radiating near field exists, and reactive and radiating terms can coexist. The labels describe which behaviour dominates and whether plane-wave assumptions are valid.
- Are most 2 m interference problems far field? Not necessarily. A nearby cable, PCB loop or hand-held transmitter can couple in the near field even at VHF, while another source may reach the same victim by conduction.
- Does common-mode current prove near-field coupling? No. Common mode describes a current relative to another return structure. It can be driven by conduction, local electric or magnetic coupling, mode conversion or an incident far field.
- Can I calculate H from E using 377 Ω? Only where a plane-wave approximation is valid. In a reactive near field, E and H must generally be considered separately.
- What is the best first RFI test? Record a repeatable receiver baseline, change one safe variable, restore it and repeat. Then pair the receiver result with a probe or current measurement that tests the suspected path.