Why I²R Matters: Where RF Current Becomes Heat

My practical rule is simple: follow the current, then ask what each part does with the power. Some power becomes wanted radiation. Some becomes heat. A low SWR, a large component or a generous power label cannot separate those outcomes for you.

Start with Current, Resistance and Power

Electric current is the rate at which electric charge passes a point. Its unit is the ampere, abbreviated A. A steady current keeps one direction. Radio-frequency current repeatedly changes magnitude and direction.

Resistance is the real part of a path’s impedance, measured in ohms, abbreviated Ω. It accounts for net average power leaving the electrical port. In an ordinary resistor that power becomes heat; in an antenna model, part of the real resistance can instead represent wanted radiation. At radio frequencies, conductors, contacts and other materials can have an effective resistance that differs from their direct-current resistance.

Power is the rate at which energy is transferred. Its unit is the watt, abbreviated W. One watt means one joule of energy per second. The instantaneous electrical power entering a two-terminal part is voltage multiplied by current:

Four loss names will recur:

• Conductor loss is heat produced by current in a conductor’s real alternating-current resistance.
• Contact loss is heat produced at a joint, switch or connector with finite and possibly unstable contact resistance.
• Ground loss is power dissipated in soil and other parts of the installed antenna return environment.
• Core loss is power dissipated in a magnetic material as its magnetic state changes; frequency, flux, waveform, temperature and material matter, so it is not automatically one constant wire resistance.

Efficiency is useful output power divided by input power, with both powers and their reference planes named. It is a ratio, not a synonym for low SWR.

Lower-case p(t), v(t) and i(t) mean values at one instant. At RF, power can move into a reactive field during part of a cycle and return during another part. The useful heating or delivered-power quantity is therefore the average of p(t) over a complete repeating interval.

RMS Turns a Changing Current into a Heating Equivalent

RMS means root mean square. For current, take every instantaneous value, square it, average those squares over a complete period, then take the square root:

T is the period being averaged. RMS is useful because a true RMS current produces the same average heating in a stated linear resistance as a direct current of the same value.

For a resistance R that is effectively constant over the waveform’s frequency content:

This is the correct home of the familiar I²R rule. For a single sinusoidal frequency, RMS voltage and current also give the real power at a port:

φ is the phase angle between voltage and current. A pure resistance has φ = 0, so cos φ = 1. An ideal reactance stores and returns energy, giving zero average power by itself.

Do not force one I²R number onto every RF material. If resistance changes with frequency, temperature or current, each frequency component needs the appropriate loss model, or the real average power must be measured directly. Ferrite-core loss, dielectric loss, radiation and nonlinear heating are not all literal wire resistors even when an equivalent resistance can represent them at one port and operating point.

The Square Term Is Why Small Resistance Can Matter

Hold the same real loss resistance constant and increase its RMS current. Doubling current produces four times the heat; multiplying current by five produces twenty-five times the heat.

RMS current	Loss resistance	Average heat
1 A	0.5 Ω	0.5 W
5 A	0.5 Ω	12.5 W
10 A	0.5 Ω	50 W

Those examples assume that the 0.5 Ω value remains real and unchanged. At operating power, a contact can warm, a conductor’s resistance can rise, a ferrite can change behaviour and an arc can turn a stable circuit into a fault. The square-law examples are a starting calculation, not a thermal rating.

Radiation Resistance Is Not Loss Resistance

An antenna feedpoint normally has a complex input impedance. The real part tells us that net average power is leaving the feedpoint, but it does not say where that power goes. In a useful single-port model, the real part can be separated into:

• Radiation resistance, Rrad: an equivalent resistance representing power carried away as electromagnetic radiation. It is not a hot physical resistor.
• Loss resistance, Rloss: an equivalent resistance representing conductor, loading, ground, joint and nearby-material dissipation referred to the same feed current.

With both quantities referred to the same RMS feed current:

Radiation efficiency is radiated power divided by power accepted at the antenna feed. The resistance ratio above is valid for that stated equivalent model and reference current. Both resistance terms can change with frequency, geometry, height, ground, return path and nearby objects.

A 50-ohm feedpoint might be mostly radiation resistance or mostly loss resistance. A vector network analyzer cannot discover that split from S11 or SWR alone. A good match is a port condition, not an efficiency certificate.

Where Real RF Loss Hides

At RF, the current distribution and material response decide the loss. Different parts need different models:

• Conductors and coils: their alternating-current resistance includes material resistivity, skin effect, proximity effect, geometry, joints and temperature. High local current density can make a small region dominate the heat.
• Contacts, switches and connectors: finite contact resistance produces heat; contamination, corrosion, loose pressure and current crowding can make it unstable. Voltage spacing and arcing are separate limits.
• Feed lines: conductor and dielectric loss depend on frequency, length, construction and temperature. Mismatch changes the voltage and current distribution, so the loss under the installed load is not determined by matched-line attenuation or SWR alone.
• Capacitors and insulation: dielectric loss and equivalent series resistance depend on frequency, material, construction and temperature. Voltage, RF current and spacing must all remain within rating.
• Ferrites and transformers: copper loss, leakage fields, parasitic capacitance and frequency-dependent complex permeability all matter. Core heating depends on material, frequency, flux, waveform, temperature and geometry; a power label or core size is not a loss measurement.
• Ground and return systems: current in conductors, connections, soil and nearby material can dissipate power. Their combined effect may be represented as loss resistance at the antenna feed, but the actual current is distributed through the installation.

A common-mode choke illustrates the boundary well. Its impedance may contain both reactive and resistive parts. The choke should reduce the unwanted exterior-cable current without exceeding voltage, flux or thermal limits. Heating is not proof that the choke is working well, and a cool choke is not proof that the current is low.

Headroom Helps, but the Label Is Not the Measurement

Large or high-power parts can provide thicker conductors, more spacing, greater surface area and more thermal mass. Those can be useful engineering margins. They do not guarantee lower insertion loss, the right ferrite material, sufficient common-mode impedance, a suitable self-resonant frequency or stable contacts.

Choose a part against the actual operating conditions:

• frequency range and complex source and load impedances;
• RMS current, peak voltage and internal branch stress;
• waveform, peak-envelope power, average power and duty cycle;
• measured loss or impedance at the intended reference planes;
• temperature rise, airflow, enclosure and ambient temperature;
• weather, contact pressure, ageing and maintenance.

QRP means relatively low transmitter power; QRO means high power. While a component remains linear, a given loss in decibels removes the same percentage at either level. High power creates more absolute heat and electrical stress. At low power, the heat may be small, but the lost link-budget margin can still matter. Neither case justifies buying by power rating alone.

Decibels Describe a Power Ratio, Not Its Cause

A decibel, abbreviated dB, expresses a ratio. If a passive stage receives net input power Pin and delivers net output power Pout at declared reference planes:

Genuine dissipative loss	Power fraction lost	Heat at 100 W input	Heat at 1 kW input
0.1 dB	2.28%	2.28 W	22.8 W
0.5 dB	10.87%	10.87 W	108.7 W
1.0 dB	20.57%	20.57 W	205.7 W
3.0 dB	49.88%	49.88 W	498.8 W

The table applies only when the dB figure is established as dissipative loss between those net-power planes. Return loss, mismatch loss, path loss and a receiver S-meter reading are different quantities. A raw S21 trace in a 50-ohm fixture can also include source, load and fixture mismatch rather than only heat inside the device.

Measure the Power Boundary You Are Claiming

Start with a drawing of the station: transmitter connector, tuner, feed line, transformer or choke, antenna feed, radiator and return system. Mark a reference plane wherever power or impedance is specified.

Then combine measurements that answer different questions:

• Complex impedance: record resistance and reactance across frequency at a calibrated plane, not SWR alone.
• Small-signal two-port data: use the required S-parameters, fixtures and actual or reproduced terminations. State calibration, de-embedding and uncertainty.
• Operating-power balance: measure net input and output power with suitable directional or power-ratio methods under the real waveform and duty cycle.
• Current and voltage: measure or calculate the relevant branch quantities, including feed-line exterior current where common mode is suspected.
• Temperature: log warm-up and steady temperature at contacts, coils, capacitors and ferrites. Temperature reveals the thermal response; converting it to watts or assigning its source requires a thermal model or calibration.
• Antenna efficiency: use a complete accepted method such as gain/directivity comparison, Wheeler-cap or reverberation-chamber measurement, with its assumptions and uncertainty.

Use an A/B/A sequence when changing one part: baseline, change, then restore the baseline. Keep frequency, power, tuner state, feed-line route, environment and measurement planes fixed. A result that appears with B and returns to its earlier value with the restored A is stronger evidence for the component effect. Failure to restore the baseline exposes drift or another confounding change.

Keep the Heat Test Safe

RF systems can produce dangerous voltage, current, contact burns, arcing and retained charge even when average transmitter power sounds modest. De-energize before changing connections. Discharge capacitors and lines, guard exposed conductors, respect component and cable ratings, and stop on arcing, smoke, unstable impedance or unexpected heating.

Do not touch a component to decide whether it is “only warm.” Use a suitable sensor with electrical isolation and a declared emissivity or contact method. Average power, duty cycle, transmission duration, airflow and ambient temperature belong in every thermal result.

The Bench Rule I Keep

I²R is powerful because it turns hidden resistance into a heat estimate. It is dangerous when used without asking which current, which resistance, which frequency and which reference plane.

Follow the current. Separate wanted radiation from real dissipation. Measure the loss at the operating condition. Then give every conductor, contact, dielectric and magnetic part enough electrical and thermal margin. That is how current becomes an efficiency result instead of folklore.

Primary and Authoritative Technical Sources

• BIPM: The International System of Units—current SI definitions and relationships for the ampere, ohm, joule and watt.
• IEEE 145-2025—current antenna impedance, accepted power, radiation resistance and efficiency terminology.
• Recommendation ITU-R BS.705-2—current HF antenna gain, efficiency, ground and practical-installation framework.
• NIST: The Interpretation and Use of S-Parameters in Lossy Lines—imperfect conductors, propagation constants and lossy-line measurement.
• NIST: Power-Ratio and VNA Attenuation Measurement—measurement technique and uncertainty boundaries for low-loss two-port devices.
• Fair-Rite: Design Considerations for High-Frequency Magnetic Materials—frequency-, material-, geometry-, temperature- and flux-dependent ferrite loss.
• IEEE 149-2021—antenna impedance, gain, pattern and efficiency measurement practice and uncertainty.

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