RF Noise in the Shack: Diagnose the Coupling Path Before Filtering

Start With Four Different Questions

A receiver display, an S-meter rise or a comb of carriers does not identify the mechanism. A useful diagnosis separates four layers:

1. Source: Which converter, clock, display, network port, motor drive, lamp or natural process generates energy?
2. Coupling path: Does energy travel through a cable, through an electric or magnetic near field, or as a radiated far field?
3. Mode and boundary: Is the measured quantity differential between two conductors, common mode on a complete cable relative to its surroundings, or current on a shield exterior?
4. Victim mechanism: Does the wanted antenna receive it, does a cable inject it into the receiver, or does strong out-of-band energy overload a front end?

One physical installation can contain all four. A switching supply can create differential ripple, drive common-mode cable current through parasitic capacitance, radiate from the resulting cable structure and overload a nearby receiver. The remedies are not interchangeable.

Radiated and Conducted Are Test Boundaries

“Radiated” and “conducted” describe how a disturbance is observed at a declared boundary. They are not permanent labels attached to a source. Current conducted onto a cable can make that cable radiate; a field can couple onto a cable and then arrive at a receiver port as a conducted signal.

The common shorthand about distance also needs care. In an unobstructed far field, electric-field magnitude is proportional to approximately 1/r and power flux density to approximately 1/r². That does not justify an inverse-square field-strength rule inside a shack. Reactive near fields, cable resonances, reflections, shielding apertures and bonding geometry can produce very different changes with distance.

Define Common Mode and Differential Mode at the Cable

For a two-conductor port, differential current flows out on one conductor and returns on the other. Common-mode current is the in-phase or net longitudinal component of the chosen conductor bundle relative to an external reference and return path. With coax, the intended transmission-line current is on the centre conductor and the shield’s inner surface; current on the shield exterior belongs to the external mode.

A clamp probe around the complete cable responds to net longitudinal current and is useful for screening common-mode current. A voltage probe from one supply conductor to its return observes a differential quantity. Neither measurement, by itself, proves what reaches the receiver input.

Protective Earth Is Not an Optional RF Component

Additional earth electrodes and floating metalwork can develop hazardous potential differences during faults, surges or lightning. Protective earthing and bonding exist for safety; their arrangement is governed by the electrical installation, not by an RF-noise experiment. Any additional electrode must be integrated into the earthing and lightning-protection design as required by the applicable rules.

• Keep required protective-earth conductors continuous and unimpeded.
• Bond additional electrodes and exposed conductive parts as required by the applicable installation and lightning-protection design.
• Do not put a common-mode choke in a protective-earth conductor.
• Do not assume that a short copper strap is an RF reference at every frequency; its inductance and the full current loop still matter.
• Treat an “RF ground” as a description of a measured return network, not permission to defeat protective bonding.

Bonding equipment cases can improve both safety and EMC when the equipment and installation are designed for it, but a new strap can also move RF current into another path. Verify the result rather than promising an automatic noise reduction.

A Repeatable Ham-Station Diagnostic Sequence

1. Freeze the receiver setup. Record frequency, mode, bandwidth, preamplifier/attenuator state, AGC state, antenna, time and a calibrated or at least repeatable level indication.
2. Establish the antenna contribution. Compare the station antenna with a screened termination at the same receiver reference plane. Protect the input and never transmit into the measurement setup.
3. Remove suspect loads safely. Use normal switches or circuit breakers and observe whether the signature changes. Battery operation can be useful only when it preserves required safety and does not introduce a different converter or ground path.
4. Change one cable at a time. Disconnect data, control, display and accessory leads only where safe and permitted. A changed noise level shows involvement, not necessarily that the disconnected device was the original generator.
5. Screen the field. A portable receiver or near-field probe can locate candidates. Probe orientation, distance and cable placement must be repeatable; a loud local reading is not a compliance measurement.
6. Measure cable current. Clamp around the whole cable, establish probe transfer impedance and receiver calibration if quantitative current is required, and scan at several positions because the external mode can form standing waves.
7. Separate DM and CM where practical. Use suitable differential probes, current-probe arrangements or a defined artificial mains network/LISN with the required isolation and safety procedures. Ordinary oscilloscope ground clips must not be attached casually to mains conductors.
8. Change one remedy. Re-run the same measurements before accepting ferrite, rerouting, shielding or filtering.

What Ferrite and Shielding Can—and Cannot—Do

A ferrite choke presents a frequency-dependent complex impedance to the mode that threads its aperture. It does not “block RF” universally. Its effect depends on material, exact part, turns, cable geometry, current, temperature, parasitic capacitance and the common-mode source/load impedances. Measure the current before and after, across every band that matters.

Shield quality is likewise a system property. Cable transfer impedance, connector termination, pigtails, seams, apertures and bonding all matter. A double shield can reduce coupling, but changing coax does not cure noise that arrives on the shield exterior or through another lead. A nominal 360-degree connector is useful only if the complete assembly maintains that current boundary.

A feedpoint choke can reduce antenna-to-feedline mode conversion, but it cannot be prescribed for every noise symptom. A second choke at the shack can change the external network; it may help, do little or move a current maximum. Current measurements decide.

Mains Filters Require Both EMC and Safety Engineering

A mains filter is not a generic station accessory to install experimentally. Use a complete, appropriately approved assembly rated for the supply system, voltage, current, prospective fault conditions, temperature and environment. Its installation must preserve protective earth, enclosure protection, creepage, clearance, strain relief and touch-safe construction.

Filter capacitors to protective earth can create leakage current; stored energy can remain after disconnection; inductors can saturate or heat; and source/load impedance can move resonances. IEC 60939-3 separates performance from safety testing for relevant passive EMI filter units. The filter’s attenuation plot under a specified fixture does not guarantee the same result in a shack installation.

Extra ferrite is not automatically redundant when a mains filter is fitted, and it is not automatically beneficial either. A certified filter may mainly address one mode or frequency range; a cable choke changes common-mode impedance elsewhere. Measure both the cable current and receiver result, then check leakage, temperature and normal equipment operation.

Compliance Is Not the Same as “Quiet in My Receiver”

CISPR product standards use defined ports, detectors, bandwidths, fixtures, distances and limits to make results reproducible. Passing an applicable emission test does not mean zero energy on every amateur frequency or guaranteed coexistence with a nearby sensitive antenna. Conversely, hearing a device at close range does not prove non-compliance.

A ham-station investigation is valuable engineering screening. A compliance conclusion requires the applicable standard, calibrated equipment, defined site and uncertainty. Keep those claims separate.

Practical Decision Rule

Remove or repair the source when possible. If that is not possible, interrupt the measured coupling path at the boundary where it exists. Choose common-mode impedance for measured common-mode current, differential filtering for measured conductor-to-conductor noise, shielding for a demonstrated field/aperture path and front-end filtering for overload. Preserve electrical safety throughout, then verify the complete station rather than declaring victory from one probe trace.

Primary Technical Sources

• ITU-R P.372-17 (2024), Radio noise — current terminology and statistical treatment for natural and man-made radio noise.
• ITU-R P.525-5 (2024), Calculation of free-space attenuation — far-field free-space relations; not a near-field shack rule.
• CISPR 16-2-1 and CISPR 16-2-3:2016+A1:2019+A2:2023 — conducted and radiated disturbance measurement methods.
• CISPR 32:2015+A1:2019 — multimedia-equipment emission requirements and reproducible test procedures.
• IEC 60364-5-54:2011+A1:2021 — earthing arrangements, protective conductors and protective bonding.
• IEC 60939-3:2024 — safety and performance requirements for relevant passive EMI filter units.
• IEC 62368-1:2023 — hazard-based safety requirements for audio/video and ICT equipment, including relevant external power supplies.
• Texas Instruments, EMI Mitigation Techniques Using the TPSM33620-Q1 (2026) — a current manufacturer treatment of differential- and common-mode conducted noise and LISN-based observation.

Questions, antenna-factor records or height trials to share? Contact RF.Guru.