RF Wattmeters: Where V²/50 Ends and Measurement Begins
RF Wattmeters: Where V²/50 Ends and Measurement Begins
The numbers on a 50-ohm watt scale come from valid power-wave arithmetic. The reading, however, comes from a directional coupler, detector, calibration and display whose limits matter as much as the formula.
The short answer is no: an inline SWR/wattmeter is not merely a voltmeter with a watt scale, and it normally does not absorb the transmitter output in a large 50-ohm resistor. It samples forward and reverse traveling waves, detects those samples and maps them to indicated power. The familiar V²/50 relationship is one part of that chain.
RF safety: use a line section, element, load, adapters and connectors rated for the frequency, power, voltage, duty cycle and mismatch. Remove RF before changing connections. The official Bird Model 43 manual warns that an energized transmission line can cause severe burns, shock or death.
First Correction: V²/R Is More Than “Just Ohm’s Law”
Ohm's law relates voltage and current. The expression P = V²/R combines that relation with the definition of real electrical power, and it needs the right voltage and load model. For an arbitrary sinusoidal load, the general phasor expression is:
P = Re{VRMSIRMS*}
For a pure resistance, current is in phase with voltage and the equation reduces to P = VRMS²/R. The same form also describes one traveling wave on an ideal line whose characteristic impedance Z0 is real. In the usual low-loss 50-ohm approximation:
P+ = |V+RMS|² / Z0 = |I+RMS|²Z0
P− = |V−RMS|² / Z0
That is why a calibrated 50-ohm meter can print watts even though its sensing circuit handles only a small sample. It is also why the formula must not be applied blindly to the total voltage at one point on a mismatched line. Complex reference impedances and significantly lossy lines require more careful power-wave definitions than the real-50-ohm approximation used here.
The Matched 100 W Result Is Correct
With a single forward wave carrying 100 W on an ideal 50-ohm line:
VRMS = √(100 × 50) = 70.71 V
IRMS = √(100 / 50) = 1.414 A
Vpeak = 100.0 V; Vp-p = 200.0 V
Those values describe a sinusoidal 100 W traveling wave. The voltage is not a universal property of “100 W RF”: it changes with reference impedance, waveform and which voltage—incident, reflected or total—is being discussed.
Mismatch: Total Voltage Is Not a One-Point Wattmeter
On a real-Z0 line, the terminal quantities are the sums and differences of the traveling waves:
V = V+ + V−
I = (V+ − V−) / Z0
Pnet = P+ − P− = Re{VI*}
Suppose an ideal meter at one reference plane reports 100 W forward and 25 W reverse. The two traveling-wave voltages are 70.71 V RMS and 35.36 V RMS. At a voltage maximum they add to 106.07 V, while the currents oppose and total 0.707 A. Their product is still 75 W. At a voltage minimum the total voltage is 35.36 V, the current is 2.121 A and the product is again 75 W.
| Same lossless-line example | Total voltage | Total current | Real power |
|---|---|---|---|
| Voltage maximum | 106.07 V RMS | 0.707 A RMS | 75 W |
| Voltage minimum | 35.36 V RMS | 2.121 A RMS | 75 W |
Squaring the voltage maximum and dividing by 50 gives 225 W; doing the same at the minimum gives 25 W. Neither is the transported real power because the local impedance is not 50 + j0 ohms at both points. The directional meter avoids that trap by estimating the separate waves.
What a Directional Meter Actually Samples
A directional bridge or coupler obtains voltage- and current-related samples and combines them so one port is mainly proportional to the forward wave and the other to the reverse wave. In the ideal real-Z0 model, the separation corresponds to combinations of V + Z0I and V − Z0I. Implementations differ—transformer bridges, transmission-line couplers and modern multiport sensors do not share one schematic—but the purpose is the same.
The official Bird Model 43 instruction book is a useful concrete example. It describes a traveling-wave line section and plug-in element, with orientation selecting forward or reflected indication. The detector output then drives the meter movement. That is directional sampling, not a direct measurement across a 50-ohm terminating resistor.
A 30 dB coupling example is useful if its assumptions are stated. An ideal, matched 30 dB coupled port receives 1/1000 of the main-line power, or 1/31.62 of its voltage-wave amplitude:
| Main-line wave power | Ideal coupled power | Coupled-port voltage into 50 Ω |
|---|---|---|
| 10 W | 10 mW | 0.707 V RMS |
| 100 W | 100 mW | 2.236 V RMS |
| 1000 W | 1 W | 7.071 V RMS |
Those are coupled-port calculations, not generic detector voltages. A practical sampler may use a different coupling factor, port impedance, termination, attenuator or detector architecture.
Coupling and Directivity Are Different Specifications
Coupling says how much of the wanted traveling wave reaches the sample port. Directivity says how well the unwanted direction is rejected. Finite directivity lets some forward signal leak into the reverse channel, where it vector-adds to the true reverse sample. The leakage has phase, so the result can read high or low.
For an illustrative coupler with 30 dB directivity, the unwanted amplitude relative to a full reflection is:
ε = 10−30/20 = 0.03162
An otherwise perfect load could therefore appear to have |Γ| ≈ 0.03162, equivalent to about 1.065:1 SWR. A true 1.20:1 load has |Γ| = 0.09091. In a simple worst-phase bound, the indicated magnitude can fall between 0.05929 and 0.12253, or roughly 1.13:1 to 1.28:1 SWR. This is an error illustration, not a complete uncertainty specification: detector errors, calibration, mismatch and frequency response still apply.
The Bird manual specifies greater than 25 dB directivity for the cited line section and explicitly recommends a more sensitive element for low reflected-power readings. Rohde & Schwarz’s directional-coupler parameter paper and Anritsu’s directivity guidance likewise treat directivity as a vector-leakage limit, not as the coupling value.
Practical implication: the reverse channel is hardest to trust when the true reflection is small. A reassuring low SWR can be limited by the sampler’s directivity and range even when the forward-power indication looks plausible.
The Watt Scale Represents a Calibration Chain
Traveling wave → directional sample → RF detector → signal conditioning/display → indicated watts
A printed scale or digital value is valid only within the sensor’s specified conditions. Important terms include:
- Frequency range and flatness: the coupling network and detector response change with frequency. The correct element or range must cover the operating frequency.
- Power range: many accuracy statements are percentages of full scale, not percentages of the reading. The Bird Model 43 manual, for example, specifies ±5% of full scale for CW. A 20 W indication on a 100 W range therefore inherits a ±5 W instrument term—±25% of that reading—before other errors.
- Detector transfer: diode detectors can traverse square-law and linear/peak-response regions. Temperature, source impedance, loading and RF/video bandwidth matter. A calibrated design compensates some of this; “a diode has a threshold” is not a sufficient meter model.
- Sensor mismatch and insertion: the wattmeter is a real line section with finite VSWR. Its interaction with source and load can change both the sampled wave and the system it is measuring.
- Display dynamics: meter movement, video filtering, sampling rate, averaging interval and peak-hold logic determine what a changing waveform produces on the display.
- Connector and handling repeatability: adapters, damaged contacts, loose interfaces and changed reference planes can dominate a careful calibration.
Bird’s figures illustrate one product family, not all SWR meters. Keysight’s official RF and microwave power-measurement fundamentals distinguishes thermal and diode sensor behavior, average and peak measurements, calibration and traceability. NIST Technical Note 1379 shows why power-sensor calibration is a transfer measurement and why source/sensor mismatch belongs in its uncertainty.
Average Power, PEP and Meter Ballistics
For a periodic waveform, true time-average real power is:
Pavg = (1/T) ∫0T v(t)i(t) dt
Across a matched, purely resistive 50-ohm load, a true-RMS voltage measured over the same interval gives Pavg = VRMS²/50. That does not make an ordinary diode-and-needle instrument a true-RMS power meter.
Peak envelope power is different. Current US 47 CFR §97.3 defines PEP from the average power during one RF cycle at the crest of the modulation envelope under normal operation. It is neither long-term average power nor the instantaneous product of one voltage and current sample.
| Signal and instrument | What a defensible reading means |
|---|---|
| Key-down CW or another constant-envelope carrier | Average and PEP during the keyed interval coincide once settled, if the sensor is specified for the frequency and level. |
| SSB voice on a basic analog meter | The indication depends on detector response and meter ballistics; it is not automatically PEP. |
| Specified peak-reading meter | PEP is credible only within the stated envelope bandwidth, crest factor, pulse width/repetition limits, time constant and accuracy. |
| Digital or pulsed waveform | Use a sensor whose average/peak mode, RF bandwidth, video bandwidth and statistics match the waveform. |
The distinction is visible even within the Bird family. The Model 43 manual describes the base instrument for CW, AM, FM and TV rather than pulsed measurements. It states that the ordinary diode indication on AM responds mainly to carrier. The Model 43P adds a specified peak mode with a one-second time constant, ±8% full-scale accuracy and documented waveform limits. “Peak-reading” is therefore an instrument specification, not a label that can be inferred from a lively needle.
Forward Minus Reverse: Power at One Reference Plane
For the real-reference, low-loss approximation, the net power crossing the meter plane toward the load is:
Pnet = Pforward − Preverse
With 100 W forward and 25 W reverse, that is 75 W at the meter plane. It is not automatically the power accepted by a distant antenna and certainly not radiated power. Feed-line loss after the meter changes the forward and reverse waves before they reach that load; tuners, traps, coils, conductors, common-mode paths and ground loss can dissipate power; the antenna converts only part of its accepted power to radiation.
Reference plane therefore matters. Move the meter, and forward/reverse values can change on a lossy line even when the load has not. This is why where SWR is measured and the distinction between transmission loss and mismatch loss are measurement questions, not word games.
How SWR Is Derived—and When the Result Is Valid
If the two channels represent forward and reverse powers at the same real reference impedance, frequency, waveform and plane:
|Γ| = √(Preverse/Pforward)
SWR = (1 + |Γ|)/(1 − |Γ|)
For 100 W forward and 11.11 W reverse, |Γ| is one third and SWR is 2.00:1. A cross-needle scale can use this ratio geometrically. The calculation does not remove instrument errors: unequal channel calibration, finite directivity, response near the bottom of a range, modulation, frequency error and sensor mismatch all propagate into SWR.
What the 50-Ohm Label Really Means
Fifty ohms is the meter’s nominal reference and line impedance, not a claim that every connected load or every point on the line is a 50-ohm resistor. The bridge geometry, internal line section, element calibration, directional null and watt conversion are designed around that reference.
A 75-ohm system needs an instrument designed and calibrated for 75 ohms. Inserting a 50-ohm section into it creates discontinuities and invalidates simple assumptions. Likewise, a load reading 50 + j0 ohms at the meter plane does not mean every voltage and current elsewhere in the system has the matched-line value.
A Defensible Check or Calibration Workflow
- Define the measurand: forward, reverse, net average power, PEP or SWR; state frequency, waveform, reference plane and averaging interval.
- Choose the correct instrument configuration: rated line section, element or range, connectors, frequency coverage, power, duty cycle and peak/average mode.
- Use an appropriate load: a rated low-reflection termination whose uncertainty is known at the test frequency and temperature.
- Establish a power reference: use a calibrated sensor and stated traceability chain, not merely a transmitter display or a second uncalibrated meter.
- Measure forward and reverse behavior separately: check the normal direction, reverse the line section or sensor only as the manual permits, and quantify the low-reflection/directivity floor.
- Exercise frequency, range and level: one agreeable midscale reading does not validate QRP, VHF, peak or full-power operation.
- Build the uncertainty budget: include calibration factor, coupling flatness, detector linearity, full-scale term, directivity, source/sensor mismatch, insertion, adapters, repeatability, temperature and display resolution.
The practical verdict: a 50-ohm watt scale is grounded in power-wave arithmetic, but the indicated watts are the output of a calibrated measurement system. Trust the reading only to the extent that the coupler, directivity, detector, range, waveform response, reference plane and uncertainty support it.
Mini-FAQ
- Is the watt scale on an SWR meter based on Ohm's law? Partly. Its 50-ohm calibration uses valid voltage, current and power relationships, but the indication also depends on directional sampling, detection, range, frequency, waveform and calibration.
- Does an inline SWR wattmeter contain a big 50-ohm resistor? Usually not. It normally passes RF through a low-loss line section and samples it. A dummy load or absorption wattmeter is the instrument that deliberately absorbs the transmitter power.
- Why is a simple RF voltage reading not enough to measure power on coax? Under mismatch, forward and reverse waves create position-dependent total voltage and current. Squaring that local voltage and dividing by 50 does not generally give the net transported power.
- How does directivity limit a low reflected-power reading? Some forward sample leaks into the reverse channel and adds with an unknown phase. Near a good match, that leakage can be comparable with the true reflection and can make the indicated SWR read high or low.
- Why do inexpensive SWR wattmeters sometimes disagree? Their coupling, directivity, frequency response, detector transfer, range, full-scale accuracy, waveform response, insertion mismatch and calibration can differ.
- Does forward power on an SWR meter equal radiated power? No. It is a traveling-wave quantity at the meter plane. Feed-line and matching losses change what reaches the antenna, and antenna efficiency determines how much accepted power is radiated.