RX Chokes: Stop Thinking Only Like a Transmitter

I understand why a transmitter-trained operator reaches for a large ferrite core. On transmit, common-mode current, ferrite heating, voltage stress and unwanted radiation can be urgent problems. Receive changes the power level and the measurement objective, but it does not create a separate kind of common mode. A correctly chosen coax choke can be useful on RX; a blindly copied one is still guesswork.

RX Changes the Objective, Not the Mode Physics

A transmitting station may choose a choke to reduce feed-line radiation, station RF, pattern distortion or ferrite heating while carrying substantial RF power. A receiving station usually cares about a different set of outcomes:

• less current on the outside of a coax shield or another unintended conductor;
• less coupling of local electrical noise into the receive port;
• less mode conversion at an unbalanced interface;
• more wanted-signal-to-noise ratio, not merely a lower S-meter reading;
• more receiver headroom in the presence of strong signals; and
• negligible change to the intended differential signal path.

Those objectives do not make a large coax choke automatically wrong, nor do they make a surface-mount common-mode component automatically right. The correct part is the one whose ports, complex impedance, parasitics, bandwidth and limits fit the diagnosed path.

Define the Conductors and the Reference

“Common mode” is incomplete unless the conductor set and reference are stated. ITU-T K.10 defines common-mode voltage and current for a specified pair relative to a reference conductor. In a coaxial installation, the wanted transmission mode uses current on the centre conductor and the facing inner surface of the outer conductor. Current on the exterior surface of the shield closes through some combination of antenna structure, mast, building wiring, bonding, soil, equipment chassis, operator capacitance and other cables.

That exterior-current loop can receive local noise, change the installed antenna response or carry RF between the antenna and station. But its voltage and current are not separate “RX” and “TX” enemies. They are linked by the complete common-mode source, path and load impedances:

This simple series expression is a planning model, not a universal 50 Ω divider. A real installation is distributed and may have several coupled or parallel return paths. It nevertheless exposes the central point: the same choke impedance can produce a large current change in one installation and almost none in another.

A Choke Adds Complex Impedance to One Path

A ferrite sleeve, bead string, wound coax core or common-mode component adds a frequency-dependent complex impedance to the conductors passing through it. The resistive part dissipates common-mode energy; the reactive part stores and returns energy. Winding capacitance, leakage, core geometry, cable geometry and nearby conductors can create resonances and change both parts across the band.

A catalogue material number is not the installed choke impedance. Fair-Rite, for example, states that its material curves are measured on specified standard specimens. Change the core dimensions, number of passes, cable, winding spacing, enclosure or temperature and the finished device changes too. A quoted impedance without frequency, resistance, reactance, fixture, conductor arrangement and uncertainty is not a complete choke specification.

Nor is “more ohms” automatically better. Added impedance must be considered with the rest of the loop, and a high-Q resonance can move sharply with installation capacitance. The useful target is enough stable common-mode impedance over the required band to reduce the measured unwanted current or transfer without creating unacceptable differential loss, mode conversion, heating or voltage stress.

Keep Differential and Common-Mode Results Separate

A coaxial choke is intended to impede exterior-shield current while allowing the wanted coaxial mode to pass. That ideal separation is finite. Connectors, winding geometry, parasitic capacitance, cable loss, asymmetry and fixture transitions can alter:

• differential insertion loss and return loss in the wanted receive path;
• common-mode transmission along the unintended path;
• common-to-differential and differential-to-common conversion; and
• amplitude and phase balance at a balanced interface.

A surface-mount common-mode choke on a balanced pair, a transformer at a port and a ferrite choke around a complete coax are therefore not interchangeable labels. Each acts on a defined conductor set and has different parasitic and voltage/current limits. Draw the ports and current paths before selecting the component.

CMRR Starts with a Defined Port and Source Balance

Common-mode rejection ratio, or CMRR, is the ratio of differential gain to common-mode gain under declared input, output, frequency, level and termination conditions. It is an amplifier or network result; it is not another name for choke impedance.

A high-CMRR differential input can still perform poorly when the two generator impedances, traces, protection parts or cable connections are unequal. That imbalance converts part of the arriving common-mode voltage into a differential signal before or at the active device. Conversely, a well-balanced source and interface may need less added impedance than an asymmetric installation. CMRR, source balance, mode conversion and common-mode path impedance must be checked together.

An active receive antenna is not automatically a megohm voltage probe or an unbalanced input. E-field probes, magnetic loops, buffered whips and balanced sensors use different interfaces. The antenna’s internal sensor impedance, amplifier input, output port, feed arrangement and power path must be taken from its actual schematic or measured ports.

Placement Follows the Current Path

“At the antenna” and “at the receiver” are both useful hypotheses, not laws. A choke can act only on the path that passes through it. The important boundary may be:

• the active antenna’s RF output;
• the point where coax leaves the antenna structure or mast;
• the building entry and bonding plane;
• the receiver or multicoupler input;
• a DC feed, control, Ethernet or USB cable; or
• more than one point when parallel paths remain.

Transformer coupling and an isolated DC supply can reduce galvanic continuity, but neither creates infinite RF isolation. Interwinding capacitance, enclosure capacitance, protection parts, cable screens and nearby conductors can bridge the boundary. Hard bonding can help EMC and safety at a defined entry plane, while an uncontrolled loop elsewhere can increase coupling. The topology and applicable electrical-safety rules decide; slogans about “breaking every ground loop” do not.

Map the exterior current along the cable, vary one route or connection at a time, and observe which change affects current, signal, noise or overload. Place suppression where the diagnosed path crosses a boundary you can control. If another cable remains a parallel return, deal with that path rather than adding ferrite blindly to the first cable.

Receive Loss and Headroom Still Matter

Loss ahead of the first effective low-noise stage reduces the available wanted signal and the external-noise margin. In a noise-dominated HF installation, a small amount of loss may be tolerable; in a quiet band, at VHF, or with an inefficient sensor, the same loss may reduce sensitivity. Differential insertion loss must therefore be measured over the wanted band rather than assumed negligible.

A quieter display is not enough. The choke may have reduced a local-noise path, changed the antenna pattern, attenuated the wanted path, altered the match, or moved the receiver out of compression. Record wanted signal and same-bandwidth noise separately. When a displayed bin contains signal plus noise, subtract noise in linear power before calculating SNR; do not subtract dB readings directly.

Also check receiver headroom. Strong broadcast, amateur or local digital signals can create blocking, desensitisation or intermodulation. A network that adds wanted-band loss may appear to “clean up” the receiver simply by acting as an attenuator. Compare it with a known attenuator and inspect strong-signal products before assigning the improvement to common-mode rejection.

Thermal and Power Limits Do Not Disappear on RX

A receive-only installation can still see energy from a nearby transmitter, coupled antenna, static discharge, surge or DC bias. Small-signal impedance data do not establish safe transmit power, pulse energy, temperature rise or insulation voltage. Ferrite loss, flux density, winding current, cable voltage, duty cycle, ambient temperature and cooling determine the thermal boundary.

Measure the Path, Then Run a Rapid A/B/A Test

A useful RX-choke test keeps the installation and receiver state under control:

1. Freeze the receive configuration. Record frequency, bandwidth, detector, preamplifier, attenuation, RF gain, AGC state, averaging, preselector and display reference. Keep the antenna, cable route, bonds and connectors fixed.
2. Map exterior current. Pass the complete coax through a calibrated RF current probe so the wanted inner-mode currents cancel in the aperture. Measure at marked positions with the same probe orientation, termination, detector and bandwidth. A single standing-wave minimum is not proof of no current elsewhere.
3. Characterize the candidate device. Measure complex common-mode impedance over the operating band in a documented fixture. For a balanced or multiport network, calibrated mixed-mode S-parameters can separate differential transmission, common-mode transmission and mode conversion.
4. Test positions, not folklore. Try the diagnosed boundary and one or more plausible alternatives. Re-scan the cable and any power or control leads. Record whether current simply moved to another path.
5. Run rapid A/B/A receive comparisons. Use a stable beacon, generator or short switching intervals. Record wanted signal, same-bandwidth noise, SNR, strong-signal products and receiver state in A, with the choke in B, and after restoring A. The restored state reveals propagation and instrument drift.
6. Check the limits. Verify differential loss, return loss, ferrite temperature and any induced voltage/current under the highest credible nearby-transmit and environmental condition.

Absolute clamp-current calibration is valuable, but a carefully controlled relative scan can still locate a path. State which result is absolute, which is relative and what uncertainty comes from probe transfer impedance, position, termination, repeatability and receiver statistics.

What Counts as an Improvement?

That is the receive mindset I want: not “small SMD good, large toroid bad,” and not “copy the transmitter choke.” Follow the current path, preserve the wanted mode, measure signal and noise separately, and keep only the suppression that survives the A/B/A test.

Primary References

• ITU-T K.10, Low frequency interference due to unbalance about earth of telecommunication equipment — conductor/reference definitions, common-mode voltage and current, and coaxial common-mode loops.
• ITU-T K.37 (01/2024), EMC mitigation techniques for telecommunication installations and systems — common-mode coupling, mode conversion, screening, filtering and bonding.
• ITU-T K.136 (11/2022), EMC requirements for radio telecommunication equipment — common-mode impedance, reference planes and conversion caused by unbalance.
• Keysight E5070B/E5071B Option TDR User’s Guide — calibrated balanced, mixed-mode, imbalance and CMRR measurement definitions.
• Fair-Rite 43 Material data sheet — specimen-specific material and impedance data, frequency range and temperature boundaries.

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