Building a 49:1 UnUn for End-Fed Random Wire Antennas

End-fed wire antennas are the simplest HF antennas you can put up. One wire, one support, feed it at one end. No radials to trip over, no center-fed hassle with coax hanging from a tree. The catch? The feedpoint impedance at the end of a random-length wire is high — typically 2000-5000 ohms depending on frequency and wire length. Your radio wants 50 ohms.

A 49:1 UnUn (unbalanced-to-unbalanced transformer) bridges that gap. 49 times 50 ohms gives you 2450 ohms on the antenna side. That's close enough to the typical end-fed impedance that your tuner can handle the remaining mismatch without breaking a sweat.

This build takes about an hour, costs $15-$30 in parts, and gives you a multi-band matching transformer that handles 100 watts without complaints.

Why 49:1 and Not 64:1 or 9:1?

The math is straightforward. An autotransformer's impedance ratio equals the square of the turns ratio. A 7:1 turns ratio gives you 7² = 49:1 impedance transformation.

Some builders use a 64:1 (8:1 turns ratio). That presents 3200 ohms to the wire. Works fine. But 49:1 has become the de facto standard because:

• It lands in the sweet spot for typical end-fed wire impedances
• A 7:1 ratio is easier to wind than 8:1 on common core sizes
• The ARRL and several QRP kit manufacturers have validated the design extensively

A 9:1 UnUn (3:1 turns ratio = 450 ohms) is a different animal entirely — that's for T2FD antennas, off-center-fed dipoles, or Beverage receiving antennas. Don't confuse the two. If you're feeding a random-length end-fed wire, you want 49:1.

Parts List

Part Specification Approx. Cost Toroid core FT-140-43 (Fair-Rite #5943002701) $5-$8 Magnet wire 18 AWG enameled copper, ~4 feet $3-$5 SO-239 connector Panel-mount UHF female $2-$3 Binding post or wing nut For antenna wire connection $1-$2 Enclosure Weatherproof junction box, ~4"×4"×2" $5-$8 Stainless hardware Bolts, ring terminals $2-$3

Core selection matters. The FT-140-43 uses Type 43 ferrite material (mix 43, nickel-zinc). This material works well from 1-30 MHz with acceptable losses. It's the right choice for HF broadband transformers.

Don't substitute Type 61 or Type 31 material without understanding the implications. Type 61 has lower permeability and you'd need more turns. Type 31 works but gets hotter at lower frequencies. Stick with Type 43 for your first build.

For QRP (under 10W), you can use the smaller FT-114-43 core. For 200W+ operation, step up to the FT-240-43. The FT-140-43 is comfortable at 100W continuous, and handles 150W with adequate duty cycle (SSB is fine, RTTY or FT8 at full power gets toasty).

Winding the Transformer

This is an autotransformer, meaning primary and secondary share a common winding. You're making a single winding with a tap.

Total turns: 14 turns Tap point: 2 turns from the bottom (ground end)

The tap divides the winding into 2 turns (50-ohm input section) and 12 turns above the tap (contributing to the full 14-turn secondary). The impedance ratio is (14/2)² = 7² = 49:1.

Wait. Let me be precise here, because I see this explained wrong in half the YouTube tutorials.

The autotransformer works by voltage division. The full winding is 14 turns. The input (coax) connects across 2 turns at the bottom. The output (antenna) connects across all 14 turns. Voltage ratio = 14:2 = 7:1. Impedance ratio = 7² = 49:1. The 2-turn section IS part of the 14-turn total — it's not 2 turns plus 14 turns.

Step by step:

1. Cut about 4 feet of 18 AWG enameled magnet wire
2. Leave a 3-inch tail and begin winding through the toroid
3. Wind 2 complete turns through the core — mark this point with a small loop or knot (this is your tap)
4. Continue winding in the same direction for 12 more turns (14 total)
5. Leave another 3-inch tail at the end

Winding tips:

• A "turn" means the wire passes through the center hole once. If the wire enters the hole, that's one turn.
• Keep turns evenly spaced around the core. Don't bunch them all on one side.
• Wind in one direction only. Reversing direction mid-winding will cancel the inductance.
• Pull each turn snug against the core. Loose windings increase leakage inductance and reduce efficiency.
• Spread the 14 turns to cover about 270-300 degrees of the core circumference, leaving a gap opposite the start.

I messed up my first 49:1 build by miscounting turns. Ended up with 15 total, which gives a (15/2)² = 56.25:1 ratio. It still worked — the tuner handled it — but the SWR was noticeably worse on some bands compared to my second, correct build. Count carefully.

Connections

Once wound, scrape the enamel coating off the wire ends and the tap loop. Use sandpaper, a knife, or a soldering iron blob to burn through the insulation. Verify continuity with a multimeter.

SO-239 connection (50-ohm side):

• Bottom of winding (start, turn 1) → SO-239 center pin
• Tap point (after turn 2) → SO-239 ground/shield

Antenna connection (high-impedance side):

• Top of winding (end, turn 14) → binding post for antenna wire
• Tap point (same as SO-239 ground) → also serves as the ground/counterpoise connection

So the tap point does double duty: it's both the coax shield connection and the counterpoise/ground connection. This is correct for an UnUn — both input and output share a common ground reference.

The Counterpoise Question

Here's the thing. An end-fed wire is not a balanced antenna. RF current has to return somewhere. Without a counterpoise, it returns on your coax shield, through your station ground, through your house wiring. You get RF in the shack, common-mode noise, and RFI complaints from the family.

Minimum counterpoise: a wire (or several) connected to the ground side of the UnUn, running in some direction away from the antenna. A quarter wavelength on your lowest operating frequency is ideal but impractical for 80 or 40 meters in most yards.

Practical approach:

• A single 17-foot counterpoise wire handles 20 meters well (quarter wave at 14 MHz = 16.8 feet)
• Add a 33-foot wire for 40-meter operation
• Lay them on the ground, along a fence, wherever they fit
• Multiple shorter radials work better than one long one

For portable operation (POTA, SOTA), I've had good results with just a 25-foot counterpoise wire thrown on the ground. Not optimal, but the SWR stays manageable and I'm not carrying a radial field up a mountain.

Testing Your Build

Before connecting an antenna, test the transformer itself:

1. DC resistance: Measure end-to-end through the winding. Should be very low — under 0.5 ohms for 14 turns of 18 AWG wire on a small core. If it's open, you have a break. If it's high, your connections are bad.

2. SWR with a dummy load: Connect a 2.7K ohm non-inductive resistor across the antenna binding posts. Connect your coax to the SO-239. Measure SWR across 1-30 MHz with an antenna analyzer (NanoVNA works great here). You should see SWR under 2:1 from roughly 3.5-30 MHz. Under 1.5:1 for most of that range is achievable with a well-wound core.

If SWR is high everywhere, check: wrong core material, miscounted turns, bad solder joints, or the connections are reversed (antenna wire on the 2-turn section instead of the 14-turn section).

3. Power test: Key up into the dummy load configuration at low power (5-10W) for 30 seconds. Touch the core. Warm is fine. Hot means something's wrong — likely wrong core material or too few turns creating insufficient inductance.

Wire Length Selection

The UnUn transforms impedance, but the actual impedance presented by the wire varies wildly with frequency. Certain wire lengths create problematic impedances (very high, very low, or highly reactive) on specific bands.

Lengths to AVOID (these create half-wave resonances that present extremely high impedance on popular bands):

• 33 feet (half wave on 20m)
• 44 feet (half wave on 15m)
• 22 feet (half wave on 10m)

Recommended lengths:

• 53 feet: Works on 80-10 meters. Avoids half-wave resonance on all major bands.
• 71 feet: Good for 80-10 meters with slightly better performance on 80/40.
• 124 feet: If you have the space. Covers 160-10 meters.
• 35.5 feet: Compact, covers 40-10 meters well.

These aren't random. They're calculated to avoid half-wave multiples on the amateur bands. An antenna analyzer at the shack end of the coax will show you exactly what the tuner sees on each band.

Weatherproofing

Mount everything in a weatherproof enclosure. The core, connections, and SO-239 need to stay dry. Water on ferrite destroys performance.

• Use a PVC junction box or a Hammond weatherproof enclosure
• Seal cable entry points with silicone or cable glands
• Apply liquid electrical tape or Plasti-Dip over the wound core as extra moisture barrier
• A drip loop in the coax below the enclosure prevents water from following the cable in

Seriously, weatherproofing is where most homebrew antenna projects fail long-term. I've seen perfectly built transformers turn into high-SWR paperweights because water wicked in through an unsealed bolt hole.

Performance Expectations

Let's be honest about what this antenna is and isn't.

It is:

• A quick, cheap multi-band antenna that works from 80-10 meters (with a tuner)
• Great for portable operations, stealth installs, or limited-space situations
• Easy to deploy: throw one end over a tree, connect the UnUn at the bottom, done
• A solid performer on 20 and 40 meters with reasonable wire lengths

It is not:

• A resonant antenna on any band (you need a tuner, always)
• As efficient as a resonant dipole or beam (expect 1-3 dB loss in the transformer + mismatch losses)
• A good antenna for 160 meters unless your wire is 120+ feet long
• Quiet on receive (end-feds pick up more noise than balanced antennas in many installations)

I'll get hate for this, but the end-fed random wire with a 49:1 UnUn is probably the most practical HF antenna for most hams, even if it's not the most efficient. A wire in the air beats a resonant dipole that's still in the box. Get it up, get on the air, and build a better antenna later when you've figured out what bands and modes you actually use.