Galvanic Isolation on RX Lines: What a 1:1 Transformer Can—and Cannot—Do

The practical idea is sound when it is stated precisely: transformation or galvanic isolation and common-mode choking are separate jobs. A receiver installation may need one, the other, both or neither. The answer comes from the current paths, transfer loss and signal-to-noise ratio in the installed system—not from the label on one magnetic component.

Start by Naming the Conductors and the Reference

On a coaxial receive line, the wanted differential or transmission-line mode has current on the centre conductor and an equal-and-opposite current on the inner surface of the shield in the ideal case. Current on the shield’s outer surface belongs to another path, completed through the antenna structure, mast, soil, building wiring, equipment chassis, protective bonding, control cables or stray capacitance.

“Common mode” is not a synonym for noise. It describes voltages or currents of a declared conductor set relative to a declared reference. ITU-T K.10, for example, defines common-mode current on a two-conductor circuit from the current sum. A common-mode current may carry interference, wanted signal energy, or both. The measurement must say which conductors the probe encloses and what return path completes the circuit.

This matters because a clamp around the whole coax responds to net enclosed current. In a well-behaved differential mode, the centre-conductor and inner-shield contributions cancel outside the cable. The remaining reading is often associated with exterior-shield current, but connector geometry, nearby conductors, probe placement and fixture loading still set the uncertainty. “The coax is noisy” is an observation to investigate, not yet a mechanism.

Galvanic Isolation Is a DC Fact and an RF Transfer Function

A transformer with two electrically separate windings has no direct conductive path between its primary and secondary. A continuity test can establish that DC fact. An autotransformer cannot: its windings share a conductor. The word unun merely says that the intended ports are unbalanced; it does not prove that the windings are separate or that galvanic isolation exists.

At RF, the two sides are never separated by an infinite impedance. Interwinding capacitance carries displacement current. Capacitance between connector shells, enclosure halves, a metal mounting plate and nearby wiring adds parallel routes. As a scale example, 10 pF has a capacitive-reactance magnitude of about 1.6 kΩ at 10 MHz. That number is not the common-mode impedance of a finished assembly: all parallel return paths and their inductance must be included.

The useful statement is therefore modest. A two-winding transformer can interrupt a conductive shield-to-chassis loop while transferring the wanted RF signal magnetically. Whether that improves reception depends on how much troublesome current used that loop, how much RF still crosses capacitively, and whether another cable or bond completes the same path.

A 1:1 Transformer Is Not Automatically a Common-Mode Choke

A common-mode choke is connected so that wanted differential currents largely cancel their magnetising effect while common-mode current sees a deliberately engineered impedance. The common-mode and differential-mode impedances are different quantities and require different test connections. TDK’s measurement guidance makes that distinction explicit.

A 1:1 isolation transformer is evaluated primarily as a wanted-mode two-port: insertion loss, return loss, bandwidth, isolation topology and impedance behaviour under the intended terminations. Its turns ratio says that the ideal impedance ratio is one. It says nothing by itself about common-mode impedance, shield transfer, balance, winding capacitance or loss.

Isolation and Choking Can Be Separate Jobs

I prefer to separate the design questions. If a DC break is genuinely useful, a two-winding transformer can provide the wanted-mode transfer. If exterior current is excessive, a separate choke can be placed where its measured impedance changes that current path without unacceptable signal loss. One component need not be forced to do both jobs.

That does not create a fixed recipe such as “choke at the antenna, isolator at the receiver.” A choke placed at a current minimum may change little. A choke at a sensitive boundary or current maximum may help, but moving it changes the electrical length and termination of the common-mode path. An isolator near the receiver may be bypassed by the receiver’s USB, Ethernet, audio, power or protective-earth wiring. A remote isolator may be bypassed by a bias-T or control cable. Map the complete installation first.

Sometimes the best change is neither magnetic component. Rerouting coax, separating it from a noise-carrying cable, improving enclosure bonding, relocating an active antenna, adding suitable preselection or removing an unintended conductive bridge may attack the coupling mechanism more directly.

Bandwidth and Loss Belong to the Actual Two-Port

Mini-Circuits’ transformer guidance identifies the main boundaries. At the low-frequency end, finite magnetising inductance loads the port. At the high-frequency end, leakage inductance, interwinding and intrawinding capacitance, conductor loss and core loss shape the response. Termination impedance matters throughout. A broadband claim made with two resistive 50-ohm ports does not automatically describe an antenna presenting a complex impedance.

Measure the assembly, not just the core:

• Wanted-mode transfer: record S21, insertion loss, phase and group-delay behaviour over the receive band with declared generator and load impedances.
• Port match: record S11 and S22; a nominal 1:1 ratio does not force either port to 50 Ω.
• Isolation path: measure transfer between the two connector-shell or enclosure references with the winding ports terminated as installed. Keep the fixture’s direct coupling below the result being measured.
• Common-mode response: use the connection appropriate to the declared common mode rather than inferring it from wanted-mode insertion loss.
• Load sensitivity: repeat with representative complex impedances, cable lengths, bias networks and receiver input states.

The received signal can be very small, yet the transformer may also see a nearby transmitter, static discharge, switching transient or DC imbalance from external circuitry. Core flux rises as frequency falls for a given winding voltage and turn count. Saturation, winding loss, dielectric stress and temperature therefore need limits from the chosen core, winding, waveform and duty cycle. No ferrite mix, turn count or enclosure size is universally correct.

Balance and Layout Can Recreate the Path

Separate windings are not enough if both coax connector shells are bolted to the same conductive enclosure without an intentional isolation boundary. Mounting hardware, shields, printed-circuit copper, electrostatic screens and test-equipment grounds can all bridge the sides. Conversely, floating metalwork can develop unwanted capacitive coupling or unsafe potentials. Draw the complete schematic and the physical current paths together.

If the antenna-side port is balanced, an unbalanced-to-unbalanced transformer does not make it balanced. The interface must then be qualified for differential-to-common-mode conversion, amplitude and phase balance, and the installed antenna’s asymmetry. A DC break should never be used as evidence that exterior current has vanished.

Bias, Control and Protection Need Their Own Current Map

A galvanically isolated RF path blocks DC by design. A remote preamplifier or active antenna therefore needs power on the appropriate side, an isolated power transfer scheme, or another deliberately designed feed. Bias-T inductors, power returns, control pairs, network cables and surge components can bridge the isolation boundary at RF even when they do not appear in the signal-line drawing.

Static bleed and surge protection are not optional afterthoughts for an outdoor conductor, but their location and ratings are site-specific. A bleed device deliberately creates a path. A surge protective device diverts transient current. Both can change the RF network, and both must be coordinated with bonding and the cable-entry architecture. ITU-T K.27 treats bonding and earthing as a protection and interference-control system, while IEC 61643-21 specifies requirements and tests for surge protective devices on telecommunications and signalling networks.

Keep this discussion on a receive-only path unless every component and fault state has been qualified for transmit use. A T/R switch can fail; a nearby transmitter can couple substantial RF; a remote amplifier can place DC on the line. The lowest voltage, current, power, temperature and transient limit in the complete assembly governs.

Use a Four-State A/B/A Test

A quieter waterfall is not enough. It may mean less interference, more insertion loss, a changed receiver gain state or a moved common-mode resonance. Record wanted signal, noise and line current separately.

• Freeze the receiver: use the same frequency, bandwidth, preamp, attenuator, AGC or manual gain, reference level, FFT settings and overload state.
• Declare the planes: identify the antenna-side port, receiver-side port, enclosure reference and every auxiliary cable attached during the test.
• Record the bypass baseline: log wanted-signal level, noise power in the same bandwidth, SNR and clamp-current readings at marked cable positions.
• Test the transformer alone: insert the two-winding transformer, account for its measured insertion loss, and repeat the records.
• Test the choke alone: restore the bypass, add the choke at one marked location, and repeat. Move it only as a new declared trial.
• Test both: combine the transformer and choke without changing routing or receiver state, then repeat the measurements.
• Return to baseline: restore the first state. Rapid A/B/A repetition limits errors from changing propagation, local activity and receiver temperature.

Use a calibrated current probe or a characterised clamp fixture where possible. Record its transfer impedance versus frequency, termination, position, orientation, loading and noise floor. If a network analyser is used, verify that its cable shields and mains earth have not silently bridged the intended isolation. Optical control or battery operation can be useful measurement techniques, but they also need documented limitations.

Read the Four Results as Mechanisms

• If the transformer lowers measured common-mode current and improves SNR after insertion loss is accounted for, a conductive boundary probably mattered—but capacitive and auxiliary paths still remain.
• If the choke helps and the transformer does not, common-mode impedance at that location was more useful than a DC break.
• If the transformer and choke each help, and together help further, the two mechanisms are addressing different parts of the path.
• If the noise display falls by the same amount as the wanted signal, the apparent improvement is probably ordinary insertion loss or gain-state change rather than better SNR.
• If reconnecting USB, Ethernet, power or control wiring removes the benefit, that cable is part of the return network and belongs in the next test.
• If none of the states changes calibrated SNR or current repeatably, stop adding ferrite and look for another coupling path or receiver limitation.

The useful conclusion is not that a 1:1 “unun” cures receive noise. It is that a properly verified two-winding transformer gives the engineer one controlled DC boundary, while a separately qualified choke gives one controlled common-mode impedance. The installation decides whether either boundary matters.

Galvanic isolation breaks a conductor. Common-mode choking impedes a mode. Only measurement tells you which one the receiving system needed.

Primary and Authoritative Technical References

• ITU-T K.10—in-force definitions of differential- and common-mode voltage/current with a declared reference circuit.
• ITU-T K.136—current EMC terminology including converted common-mode current for radio telecommunication equipment.
• Mini-Circuits AN20-001: How RF Transformers Work and How They Are Measured—DC-isolated and autotransformer topologies, insertion loss, terminations, magnetising inductance, leakage inductance, capacitance, core loss and test methods.
• TDK: Measuring Common-Mode and Differential-Mode Choke Impedance—distinct fixture connections for the two modal impedances.
• TDK: Common-Mode, Differential-Mode and Characteristic Impedance—why the three impedance quantities are not interchangeable.
• ITU-T K.27—in-force bonding and earthing configurations for equipment protection and interference control.
• IEC 61643-21:2025—requirements and test methods for surge protective devices connected to telecommunications and signalling networks.
• IEC 62368-1:2023—energy-based safety requirements and safeguards for audio/video and information-technology equipment.

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