Build an HF Receive Loop Antenna — Complete Construction Guide
A well-built receive loop can transform your HF listening experience. Where a full-size transmit antenna picks up every nearby noise source equally, a shielded receive loop is much less sensitive to electric-field interference — QRM from switch-mode power supplies, LED drivers, and solar inverters — while still responding well to the magnetic component of distant radio signals. This guide builds a practical shielded loop with the same area as a 1-metre circle, covering 1–30 MHz reception with a simple JFET buffer powered over the coax.
Why a Dedicated Receive Loop?
Modern suburban and urban amateur radio environments are plagued by locally generated electrical noise — switch-mode power supplies in computers and phone chargers, LED driver circuits in lighting, solar panel inverters, and plasma televisions. Close to the antenna, much of this noise arrives as electric-field (E-field) energy. A standard wire antenna or vertical responds to both the electric and magnetic components of a radio wave, so it cannot discriminate between wanted distant signals and unwanted local noise sources.
A small shielded loop responds mainly to the magnetic component (H-field) of the wave. Local noise sources are in the antenna's near field, where the E-field usually dominates. Surrounding the loop conductor with an electrostatic shield, broken at one point so the shield does not become a shorted turn, makes the antenna much less sensitive to that E-field noise while keeping its response to the H-field of distant sky-wave signals. In noisy urban locations the practical result can be a 10–20 dB improvement in signal-to-noise ratio on 40 m, 80 m, and 160 m.
Electrostatic Shielding Effect
In this design the copper tube is the shield and an insulated wire threaded through it is the loop. E-field pickup lands on the tube and is carried to ground at the feed point instead of reaching the inner wire. The gap in the tube at the top stops the tube acting as a shorted turn; without it, current in the tube would cancel the magnetic field inside it and the inner wire would receive almost nothing. Much of the real-world benefit also comes from keeping the loop well balanced and stopping noise currents travelling up the outside of the feedline — see the common-mode choke section below.
Figure-8 Radiation Pattern
A small loop responds most strongly to signals arriving edge-on, in the plane of the loop, and has two deep nulls broadside, perpendicular to that plane. Turning the loop so its flat face points at a local noise source can reduce that noise by 20 dB or more for ground-wave and nearby sources. Nulls on sky-wave signals are shallower because those signals arrive from high angles.
Receive Only — Not for Transmit
This design is for receive use only. The small loop has extremely low radiation resistance and high reactance, making it a very inefficient transmitter, and the buffer would be destroyed by transmit power. The JFET buffer at the feed point presents a high impedance to the loop and drives the 50 Ω coax run to the receiver.
Design Overview & Theory
The antenna has three subsystems: the shielded loop element, a buffer amplifier at the feed point, and a bias tee that sends 12 V up the coaxial feedline to power the buffer. The loop is a single-turn inductor: an insulated wire inside a 22 mm copper tube shield (or the centre conductor of a length of RG-213, whose braid acts as the shield). The shield is broken at the top; the inner conductor runs continuously from one feed terminal to the other.
A 1 m diameter circle has a circumference of 3.14 m and an area of 0.785 m². The square version in this guide has 886 mm sides, the same area, and a perimeter of 3.54 m. At 14 MHz that perimeter is about 0.17λ; at 3.5 MHz it is about 0.04λ. The loop is electrically small on the lower bands, so its radiation resistance is tiny (milliohms) and its reactance is high. The open-circuit voltage across the feed terminals is proportional to the loop area and the frequency of the incident wave, so the output rises as frequency goes up.
Toward 30 MHz the loop is no longer electrically small. Its inductance, together with the capacitance between the inner wire and the shield and the buffer's input capacitance, forms a self-resonance somewhere in the upper HF range. With a high-impedance buffer like this one, expect the response to rise with frequency and peak near that resonance. That is normal; use the receiver's RF gain or attenuator on the higher bands.
Voc = 2π × f × μ₀ × N × A × H
where N = turns (1), A = loop area (m²), f = frequency (Hz), H = magnetic field strength (A/m), μ₀ = 4π × 10⁻⁷ H/m
L (µH) ≈ 1.257 × r × (ln(8r/a) − 2) — r = loop radius (m), a = conductor radius (m)
For a 1 m diameter loop (r = 0.5 m) made from 22 mm tube (a = 11 mm): L ≈ 1.257 × 0.5 × (ln(8 × 0.5 / 0.011) − 2) ≈ 1.257 × 0.5 × (ln(364) − 2) ≈ 1.257 × 0.5 × 3.90 ≈ 2.45 µH. The equal-area 886 mm square works out at about 2.56 µH. At 14 MHz the reactance is 2π × 14 × 2.45 ≈ 215 Ω, which is why the loop needs a high-impedance buffer rather than a direct connection to 50 Ω coax.
Materials required — copper tube shielded loop with JFET buffer and bias tee
JFET Buffer Circuit
The buffer is the heart of the receive loop system. The loop's output is a small voltage behind an inductive source impedance of roughly 30–300 Ω across HF, and a direct connection to 50 Ω coax would load it heavily. A JFET input stage presents a very high impedance to the loop, and an emitter follower after it drives the coax. The circuit is a buffer, not an amplifier: overall voltage gain is about −2 to −3 dB relative to the loop's open-circuit voltage. Modern receivers have enough sensitivity for this on most bands; on 160 m and 80 m in a quiet location you may want the receiver's preamp switched on.
The full schematic, parts list and assembly steps are below the build image.
Component selection notes
- Q1 — J310 JFETLow-noise, high-transconductance JFET that sets the high input impedance. It self-biases through R2 (1 kΩ); the gate sits at 0 V because the loop's inner wire and R1 connect it to ground at DC. J309 and U310 are good substitutes. Lower-current JFETs such as the MPF102, 2N5484 JFET or BF245 often bias the source too low to run Q2 properly, so stick with the J310 family. TO-92 pinouts vary between manufacturers — check the datasheet.
- Q2 — 2N3904 emitter followerGives a low output impedance that can drive 50 Ω coax with very little loss. A 2N5109 runs at the same bias and handles strong broadcast signals better.
- R4 and R6 — 47 Ω stoppersSmall resistors right at the JFET gate and transistor base that stop the followers oscillating at VHF. Mount them directly on the device pins.
- D1 and D2 — input protectionAntiparallel 1N4148 diodes limit static and large voltages from a nearby transmitter. They can generate intermodulation on extremely strong signals; keep them if you transmit from the same property.
- L1, R3, C2, C3 — phantom powerL1 takes the 12 V off the coax centre conductor without loading the RF signal; R3, C2 and C3 filter it into a clean rail of about 11 V. C1 passes RF to the coax while keeping the 12 V off Q2's emitter.
Receive Loop Inductance & Impedance Calculator
| Ref | Part | Notes |
|---|---|---|
| Q1 | J310 N-channel JFET | J309 or U310 also work |
| Q2 | 2N3904 NPN | 2N5109 for better strong-signal handling |
| D1, D2 | 1N4148 | Antiparallel input protection |
| R1 | 10 kΩ | Holds the gate at 0 V if the loop is disconnected |
| R2 | 1 kΩ | Q1 source resistor; sets the bias |
| R3 | 47 Ω | Supply filter |
| R4, R6 | 47 Ω | Gate and base stoppers; mount right at the pins |
| R5 | 150 Ω | Q2 emitter resistor |
| C1 | 100 nF ceramic | Output coupling; blocks DC from Q2 |
| C2, C3 | 100 nF ceramic, 10 µF 25 V electrolytic | Rail decoupling; watch C3 polarity |
| L1 | 100 µH RF choke, ≥100 mA | Picks the 12 V off the coax centre |
| J1 | 2-way terminal or solder pads | Inner wire ends; J1-2 also takes the tube bonding wires |
| J2 | SO-239 or BNC | Feedline: RF out, 12 V in |
| C4 | 100 nF ceramic | Bias tee DC block to the receiver |
| L2 | 100 µH RF choke, ≥100 mA | Injects 12 V onto the coax |
| C5, C6 | 100 nF ceramic, 10 µF 25 V electrolytic | Bias tee supply decoupling |
| D3 | 1N4007 | Reverse-polarity protection |
| R7, LED1 | 1 kΩ + 3 mm LED | Power indicator (optional) |
| FB | 43-mix ferrite bead(s) | On the 12 V lead into the bias tee |
| J3, J4 | SO-239 or BNC | J3 to the loop feedline, J4 to the receiver |
Buffer and Bias Tee — Assembly Sequence
Build and bench-test the electronics before you fit them into the loop. Allow 1–2 hours.
Plan the board layout
Lay the preamp out on a 50 × 40 mm piece of stripboard in the same order as the schematic: input terminals and D1/D2 at one end, then Q1, Q2, and the output and power components at the other end. Run a ground strip along one edge and connect every ground symbol to it. Keep the input and output at opposite ends so the output cannot couple back into the high-impedance input.
Fit the passive parts
Solder R1, R2, R3 and R5, then C1, C2 and C3, then L1. C3 is an electrolytic: its negative lead (the striped side) goes to ground. Keep the leads of the 100 nF capacitors as short as possible; they must work up to 30 MHz.
Fit Q1, Q2 and the stopper resistors
Check the pinout of your actual J310 and 2N3904 on the manufacturer's datasheet; TO-92 pinouts are not standard between makers. Solder R4 directly to the J310 gate pin and R6 directly to the 2N3904 base pin, with the device bodies close to the board. Q1's drain and Q2's collector both go to the +11 V rail.
Fit the input protection diodes
Connect D1 and D2 from the input node to ground facing opposite ways — one diode's band toward ground, the other's band toward the input. Together they clip any voltage above about ±0.6 V before it reaches the JFET gate.
Build the bias tee
In a small box (metal is best), connect C4 in series between the centre pins of J3 and J4. Connect L2 from the J3 centre pin to the DC node, and C5 and C6 from the DC node to ground. Wire the 12 V socket through the ferrite bead and D3 (band toward the DC node), and add R7 and LED1 from the DC node to ground as a power indicator. Join all the grounds, including both coax socket bodies.
Bench test
Connect the preamp's J2 to the bias tee's J3 with a short coax lead and apply 12 V through a multimeter set to current. Expect 10–35 mA; more than 50 mA means a wiring fault, so switch off and check. Then measure to ground: the rail should read about 10.5–11.5 V, Q1's source 1.5–4.5 V (it depends on the individual J310), and Q2's emitter about 0.65 V below Q1's source. Confirm there is 0 V DC on J4's centre pin before connecting the receiver.
Listening test and weatherproofing
Connect the receiver to J4 and touch a short wire to the preamp input: the noise level should rise clearly. Once the loop is connected, turn it to null a known local noise source. Before outdoor use, coat the board with conformal coating or a thin layer of silicone, seal every enclosure entry, and drill a 2 mm drain hole at the lowest corner of the enclosure so condensation can escape.
Copper Tube Loop — Construction Sequence
Build the copper-tube shield, thread the inner loop wire, make the shield gap, fit the JFET buffer, then mount and feed the antenna. Allow 4–6 hours for the complete build.
Choose loop geometry: circular vs. square
A circular loop gives the most area for a given perimeter. In practice a square loop with four 90° corners, built from copper plumbing tube and elbows, is far easier to make. The loop's output depends on its area, so match the area of the 1 m circle (0.785 m²): use four sides of 886 mm, measured centre-to-centre, for a perimeter of 3.54 m. At equal area the difference between circle and square is negligible — less than 0.1 dB. A 785 mm square only matches the circle's perimeter; it has 21% less area and gives about 2 dB less signal.
Cut the tube and dry-fit the frame
Each side's tube length is 886 mm minus twice the distance from an elbow's centre line to the tube end when it is pushed home — about 845 mm with typical 22 mm elbows, but measure your fittings. Cut with a pipe cutter for clean square ends and deburr the inside of every cut so it cannot damage the inner wire's insulation. Cut the top side into two equal pieces, leaving a 25 mm gap between them. Cut the bottom side into two pieces so their ends enter the preamp enclosure's end walls by about 20 mm each. Dry-fit the frame on the floor and check that it is square.
Thread the inner wire
Before soldering any joint, feed one continuous 3.8 m length of insulated copper wire through every tube section, elbow and the gap sleeve, so that both ends come out of the two bottom tube ends. Leave about 150 mm spare at each end for the connections inside the enclosure. If you use solder fittings, pull the wire back from each joint or protect it while heating, and let the joint cool before moving on. Then solder (or push-fit) the four corner joints.
Create the electrostatic shield gap at the top
The shield gap is a 20–30 mm break in the copper tube at the top, diametrically opposite the feed point. Bridge it mechanically with a plastic push-fit coupler or a plastic sleeve epoxied over both tube ends. The inner wire passes straight through the sleeve and is not cut. The gap stops circulating currents in the tube; without it, the tube would be a shorted turn around the inner wire and would block almost all of the H-field signal. Check with a multimeter: the inner wire should read under 1 Ω end to end and open circuit to the tube, and the two halves of the tube should not be connected to each other across the gap.
Mount the preamp enclosure at the feed point
Drill one 22 mm hole in each end wall of the ABS enclosure and push the bottom tube ends in by about 20 mm. Seal both entries with epoxy or silicone. Inside, fit an earth clamp to each tube end and bond both clamps to the preamp ground with short, heavy wire — the two tube halves are the shield, and both must be grounded at the feed point. Connect one end of the inner wire to the preamp input (J1-1) and the other end to the preamp ground (J1-2). Mount the SO-239 coax connector on the bottom face of the enclosure.
Mount on a non-conductive mast
The loop must be mounted on a non-conductive mast or support — PVC pipe, fibreglass rod, or wood. A conductive mast within 500 mm of the loop will distort its pattern and spoil its nulls. A 1.5–2 m length of 40 mm PVC conduit makes an excellent mast, holding the loop at head height for rotation; running it up the middle to the top sleeve also supports the shield gap. Fix the loop to the mast with 22 mm plastic P-clips. For permanent outdoor installation, solder all copper joints (not just push-fit) and seal the enclosure with marine-grade silicone.
Run the coaxial feedline and DC power
Run RG-58 or RG-213 coaxial cable from the SO-239 on the preamp enclosure to the bias tee next to the receiver. The bias tee puts 12 V DC onto the coax centre conductor, and L1 in the preamp takes it off again, so a single cable carries both signal and power. If you prefer a separate shielded DC cable, feed the 12 V through the ferrite beads straight to the preamp rail and leave L1 out. Wind a common-mode choke just below the enclosure, secure the coax to the mast with UV-resistant cable ties, and leave a drip loop so water cannot run down the coax into the connector.
RG-213 Coax Cable Loop — Construction Sequence
A cheaper and quicker alternative to the copper tube loop is to use a length of RG-213 as the shielded loop element. It is wired exactly like the tube version: the braid is the shield, broken at the top, and the centre conductor is the loop, running continuously from one feed terminal to the other. At the feed point both braids are grounded, one end of the centre conductor goes to ground and the other goes to the buffer input.
Cut 3.3 m of RG-213 and make the shield break
Cut a 3.3 m length of RG-213. This forms a circle about 1 m in diameter with enough left over for the connections. At the midpoint (1.65 m from each end), carefully cut through the outer jacket and braid only — do not cut the dielectric or centre conductor. Remove a 20–30 mm section of jacket and braid, making sure no braid strands bridge the gap. Wrap the exposed dielectric with self-amalgamating tape to keep water out.
Connect at the feed point
At each end, strip back the jacket and form the braid into a pigtail, then strip about 10 mm of dielectric from the centre conductor. Bond both braid pigtails to the preamp ground (J1-2). Connect one centre conductor to ground as well, and the other centre conductor to the preamp input (J1-1). The braids must not carry the signal: they are the shield, and both are grounded at the feed point.
Form the loop shape
Shape the RG-213 into a circle about 1 m in diameter and secure it with a few cable ties to a lightweight fibreglass or PVC support frame, with the shield gap at the top. The coax's natural stiffness holds a rough circular shape; a hula-hoop-size PVC ring makes a neater support. The feed point is at the bottom, where both ends enter the preamp enclosure.
Loop Orientation & Null Steering
The loop's figure-8 pattern has two deep nulls broadside to the loop — perpendicular to the plane of the loop. For ground-wave and local noise sources these nulls are often 20–40 dB deep on a well-built, well-balanced loop. Rotating the loop so its flat face points at a nearby noise source — a neighbour's solar inverter, a street light controller, a plasma TV — can dramatically improve the signal-to-noise ratio on affected frequencies. Mount the loop on a rotating base (a lazy-Susan bearing on the PVC mast works well) for easy null steering.
Multiple Turn Loops for Lower Frequencies
For more output on 160 m and below, a multi-turn loop (3–5 turns) raises the output voltage by the number of turns N. However, the inductance rises by roughly N², and the extra inter-winding capacitance pulls the self-resonance down, so multi-turn loops become difficult to use above about 10 MHz. For a broadband 1–30 MHz receive loop, a single turn is the best compromise.
Reducing Common-Mode Noise
Even a well-shielded loop can pick up common-mode noise that travels up the outer surface of the feedline coax. A choke balun immediately below the preamp enclosure suppresses this path effectively: about 10 turns of RG-58 on an FT-240-31 toroid, or as many turns of RG-213 as will fit (typically 6–7). A second choke at the shack end helps if noise is coming in from the house wiring. This single step can recover 10–15 dB of noise rejection that would otherwise be lost to coax shield currents.
| Band | Frequency | Loop output | Noise improvement vs wire (urban) | Notes |
|---|---|---|---|---|
| 160 m | 1.8–2.0 MHz | Lowest | 10–20 dB typical | Most dramatic noise improvement; the receiver's preamp may help |
| 80 m | 3.5–4.0 MHz | Low | 8–15 dB typical | Often transforms 80 m from unusable to workable in urban areas |
| 40 m | 7.0–7.3 MHz | Moderate | 5–12 dB typical | Strong improvement with good sensitivity |
| 30 m | 10.1 MHz | Good | 3–8 dB typical | Output keeps rising with frequency |
| 20 m | 14 MHz | Good | 2–6 dB typical | Good performance; sky noise begins to dominate |
| 15–10 m | 21–28 MHz | Good to high | 1–4 dB typical | Response may peak near the loop's self-resonance; atmospheric noise dominates |
Figures are typical of reports from urban locations and vary widely with the local noise environment. Sky noise floor: above 14–20 MHz, sky noise (galactic background and atmospheric noise) increasingly sets the noise floor rather than local man-made interference. Above this crossover frequency, the receive loop's noise reduction advantage diminishes because the noise floor is already set by the propagation medium rather than local sources. Below 10 MHz in urban areas, the receive loop is often transformative.
Can I transmit through this receive loop?
No — this design is strictly receive-only. Never connect the loop, the preamp or the bias tee to a transmitter output; even a few watts will destroy the buffer. The loop itself has extremely low radiation resistance (milliohms) and high reactance, making it a very inefficient transmitter anyway. Use the receiver or transceiver's separate RX antenna input, or a relay that disconnects and grounds the loop during transmit. D1 and D2 only protect against moderate pickup from a nearby transmit antenna.
How does this compare to a commercially built loop like the Wellbrook ALA1530?
Commercial loops such as the Wellbrook ALA1530 use carefully designed balanced, low-impedance amplifiers that give a flatter frequency response and better strong-signal handling than this simple high-impedance buffer. A home-built loop with this buffer still delivers most of the noise-reduction benefit for listening on noisy bands, at a fraction of the cost, and lets you choose the loop size.
Does the loop need to be exactly 1 metre in diameter?
No — loop size is a practical trade-off. Larger loops collect more signal (output scales with area) but are physically harder to manage and rotate, and their self-resonance falls lower in HF. A 500 mm loop is easier to handle indoors; a 1.5 m loop gives more output on 160 m and 80 m. For most HF amateur use, 800 mm to 1.2 m across is the practical optimum. The buffer works the same across this range.
Can I use aluminium tube instead of copper?
Yes — aluminium tube is lighter and cheaper than copper and works well as the shield. Aluminium cannot be soldered with standard solder, so make the corners with aluminium or stainless bolted brackets, and use stainless hardware throughout. Do not join aluminium directly to copper fittings: the two metals corrode each other quickly outdoors. If you use anodised tube, remove the anodising where you make electrical connections, because it is an insulator. The electrical performance difference between copper and aluminium for a receive-only loop is negligible.
What JFET should I use if the J310 is unavailable?
The J309 and U310 are the best substitutes, and the surface-mount MMBFJ310 is the same die if you are building on a PCB. Lower-current JFETs such as the MPF102, 2N5484, BF245 and BF256 have a much lower drain current and pinch-off voltage, so with this circuit their source often sits too low to bias Q2 properly. Check the pinout on the datasheet whichever part you use, because TO-92 JFET pinouts vary. Bipolar transistors are used successfully in low-impedance loop amplifiers, but for this high-impedance input stage a JFET is the right choice.
How do I use the directional nulls effectively?
Identify the noise source direction if possible — a nearby street, a neighbour's house, or a known industrial location. The nulls are broadside to the loop, so turn the loop until its flat face points at the noise source. In practice, slowly rotate the loop while listening to a quiet spot on the band and find the orientation that minimises the noise. A 20–40 dB null on a local noise source can transform a noisy band into a usable one.