Homebrew SWR / Directional Wattmeter (Bruene Bridge)
Nearly every build guide in this series tells you to verify the result with an SWR check — a homebrew directional wattmeter, built around the classic Bruene bridge design, gives you real-time forward and reflected power readings while transmitting, which a NanoVNA (better suited to bench sweeps and antenna analysis) isn't designed for. This guide covers how the bridge's current transformer and voltage sampling network separate forward from reflected power, the calibration procedure every build actually needs, and an SWR-from-power-readings calculator for using the finished instrument.
What a directional wattmeter measures
A simple in-line power meter only tells you the total RF present on the line, which is the sum of forward (outgoing) and reflected (returning) power and doesn't by itself tell you anything about match quality. A directional wattmeter separates the two, reporting forward and reflected power independently — from which SWR, return loss, and how much power is actually reaching the antenna can all be derived. This directional separation is exactly what makes an SWR meter useful for real-time tuning feedback while transmitting, something a NanoVNA (which measures with the transmitter disconnected, at its own low test power) isn't built to do.
The Bruene bridge principle
Named for Warren Bruene's original design, this bridge combines two independent samples of the line's RF: a current sample, taken via a small toroidal current transformer with the coax center conductor passing through it as a one-turn primary, and a voltage sample, taken via a resistive or capacitive divider across the line. These two samples are combined in two different ways — added in one combination, subtracted in the other — so that one combination cancels for a perfectly matched load (reporting near-zero reflected power) while the other tracks the actual forward power. The elegance of the design is that both forward and reflected readings come from the same two underlying samples, just combined differently.
Why this page won't hand you a "solve for your own values" formula
Unlike the matching-network calculators elsewhere in this series, a Bruene bridge's transformer turns count and sampling resistor/capacitor values aren't something a builder typically solves fresh from a target impedance — they interact with each other in ways that published reference designs (going back to Bruene's original QST article and repeated across decades of amateur projects) have already worked out and refined. The honest, reliable path is to build to a known-good reference component set, then use the bridge's built-in calibration trimpot to null it against a real dummy load, rather than attempting to calculate exact values from scratch.
Diode detection and meter scaling
- Rectifier diodes: the sampled RF is rectified to DC by a small diode on each of the forward and reflected sampling paths — germanium (1N34A-style) or low-forward-voltage Schottky diodes are commonly favored over standard silicon diodes because their lower forward voltage drop gives better sensitivity at low RF power levels.
- Meter nonlinearity: because the diode's forward voltage isn't perfectly proportional to power across a wide range, a simple analog power meter reads noticeably nonlinear at the low end — this is a known, expected characteristic of simple diode-detector wattmeters, not a build defect, and is why many designs use range switching or a deliberately non-linear meter scale rather than expecting one linear scale to cover both QRP and full-power readings accurately.
| Component | Typical Reference Value | Notes |
|---|---|---|
| Current transformer core | Small ferrite toroid (e.g. FT-50-43 class) | See the Core Selection Guide for mix/size selection; the coax center conductor forms the one-turn primary |
| Transformer secondary | Commonly in the 30-40 turn range in published designs | Exact turns count varies by specific published design and target power level — follow your chosen reference schematic rather than guessing |
| Voltage-sampling resistors | Small ohm-range values, matched pair | Paired with the calibration trimpot to balance the bridge; exact value is design-specific |
| Rectifier diodes | Germanium (1N34A-style) or low-drop Schottky, matched pair | Lower forward voltage drop than silicon improves low-power sensitivity |
| Calibration trimpot | Small panel or board-mount trimpot | Adjusted during calibration to null reflected reading into a known-good dummy load |
| Meter movement | Basic analog movement (e.g. 0-1mA or 0-100µA range) | Full-scale calibrated to your target power range with a series resistor |
SWR from Forward/Reflected Power
Materials for a homebrew directional wattmeter
Inside a homebrew Bruene bridge directional wattmeter, showing the current transformer, sampling network, and calibration trimpot.
Building a Bruene Bridge Directional Wattmeter
Follow a proven published reference schematic for exact component values — this sequence covers the general assembly and, critically, the calibration procedure every build needs regardless of which specific reference design you follow.
Wind the current transformer
Wind the secondary onto your chosen toroid per your reference design's turns count, keeping turns even and snug. The coax center conductor (or a short bus wire carrying the through-line current) forms the one-turn primary passing through the core's center.
Build the voltage sampling network
Wire the sampling resistors and calibration trimpot per your reference schematic, keeping the forward and reflected paths' components as closely matched as practical — mismatched components here directly limit how well the bridge can null out reflected power on a matched load.
Add the diode detectors and meter circuit
Wire the rectifier diodes and any filtering components on each sampling path, then connect the meter (with its switch, if using a single meter for both readings, or as a dual-meter arrangement). Keep the two detection paths physically separated inside the enclosure where practical.
Assemble in a shielded enclosure
Mount the completed bridge in a metal enclosure with good shielding between sections — an unshielded or poorly shielded build allows stray coupling between the forward and reflected paths that directly degrades directivity, showing up as a reflected reading that never drops to near-zero even into a perfect dummy load.
Calibrate against a known-good dummy load
Connect the bridge between your transmitter and an accurately rated dummy load, apply low power, and adjust the calibration trimpot until the reflected reading drops to its minimum (ideally near zero, since a perfect dummy load presents essentially no reflection). Then set the forward meter's full-scale calibration against your transmitter's known output power. Verify at a couple of different power levels within the bridge's rated range before relying on it.
| Symptom | Most likely cause | Diagnosis | Fix |
|---|---|---|---|
| Reflected reading never drops near zero, even into a known-good dummy load | Poor directivity from inadequate shielding/layout, mismatched sampling components, or the calibration trimpot not properly adjusted | Confirm the dummy load itself is accurate first (cross-check with a NanoVNA if available), then check shielding and component matching | Improve internal shielding and separation between the forward/reflected paths, re-match components if possible, and re-run calibration |
| Both forward and reflected meters read unexpectedly high simultaneously | Wiring error, often a reversed connection on the current transformer or a swapped forward/reflected path | Compare your wiring against the reference schematic point by point, paying particular attention to transformer winding direction/polarity | Correct the wiring per the schematic and re-test |
| Meter needle pins or behaves erratically | Drive power exceeds the bridge's designed range, or a diode/meter movement fault | Reduce power and see if normal behavior returns; check diode orientation and meter movement continuity if the issue persists at low power | Operate within the bridge's rated power range, or repair/replace the faulty component |
| Low-power (QRP) readings are inaccurate or pinned near zero | Diode forward-voltage nonlinearity at low signal levels — a known characteristic of simple diode-detector wattmeters, not necessarily a defect | Compare behavior across your full intended power range rather than judging accuracy from low-power readings alone | Accept the reduced accuracy at very low power as an inherent limitation, or add a switched sensitivity range if your reference design supports it |
| Bridge's SWR reading disagrees noticeably with a NanoVNA measurement at the same point | Calibration drift, component tolerance, or simply the expected accuracy difference between a simple diode-detector bridge and a NanoVNA-class instrument | Re-run the dummy-load calibration procedure and compare again | Re-calibrate if drifted; otherwise treat the two instruments as complementary rather than expecting identical precision from both |
How accurate is a homebrew Bruene bridge compared to a NanoVNA?
Less precise, especially at low power, due to diode detector nonlinearity and the practical limits of homebrew component matching and shielding. That said, it's perfectly adequate for real-time tuning feedback while transmitting, which is a genuinely different job than a NanoVNA's detailed bench sweep — the two instruments complement rather than replace each other.
Why does shielding matter so much for this build?
The bridge's whole job is separating two very similar signals (forward and reflected samples) that differ only in a subtle combination of the current and voltage samples. Stray coupling from poor shielding lets a little bit of the "wrong" signal leak into each path, directly limiting how cleanly the bridge can null reflected power on a matched load — this shows up as directivity that never gets very good no matter how well the components themselves are matched.
Can this handle high power?
Yes, with appropriately rated components — the current transformer core and sampling resistors both need to be sized for your target power level, following the same power-sizing logic as the Core Selection Guide. A design built and calibrated for 100W isn't automatically safe or accurate at the legal limit without reviewing component ratings for the higher power level.
Is this a substitute for a NanoVNA?
No — they serve different purposes. A NanoVNA is better suited to detailed antenna sweeps, impedance analysis, and bench work where you want precision across a frequency range. A Bruene bridge wattmeter is better suited to real-time, in-line monitoring while actually transmitting, which is exactly the situation a NanoVNA isn't designed for. Most well-equipped stations benefit from having both.