Single-band Dipoles

Half-wave, folded, inverted-V, sloper, vertical dipole — the canonical resonant antenna; DIY build and commercial-buy options for a clean single-band performer

Section	Topic
1	About this volume
2	Geometry & theory of operation — the half-wave dipole
3	Feedpoint impedance and matching
4	Radiation pattern — azimuth, elevation, polarization
5	Frequency response & SWR curve
6	Variants — folded, inverted-V, sloper, vertical dipole
7	Best-case use
8	Worst-case use
9	Power handling
10	DIY build — a 40 m half-wave dipole, step by step
11	Commercial buys — current good-quality manufactured options
12	Companion gear
13	Common gotchas and myths
14	Resources

1. About this volume

The half-wave dipole is the antenna every other antenna is compared to. Decibels referenced to a dipole carry the suffix dBd, and the conversion dBi = dBd + 2.15 is the linguistic glue between free-space gain and what the rest of the literature actually uses. If you understand a dipole — its feedpoint behavior, its radiation pattern, how ground proximity rearranges its lobes, and what trade-offs the obvious variants impose — you understand most of the practical antennas in this series.

This volume is single-band first. The multi-band relatives (off-center-fed, fan, trap, doublet, G5RV, ZS6BKW, linked) live in Vol 7; they are dipoles in spirit, but the multi-band problem reshapes them enough that mixing the treatments would muddy both. The half-wave dipole here is the reference — clean geometry, clean math, clean pattern. The variants in §6 are the modest mechanical and topological tweaks that still belong to the single-band dipole family (folded, inverted-V, sloper, vertical dipole) — they bend the pattern, the impedance, or the deployment without leaving single-band territory.

The 2.15 dBi free-space gain of a dipole, the ground-effect lobing that lifts a horizontal dipole to ~7.8 dBi at h = λ/2, the 73 + j42 Ω feedpoint that trims to 73 Ω resistive once you remove the end-effect — these are the numbers a deep dive on antennas builds on, and they’re the numbers this volume nails down before the rest of the series leans on them.

2. Geometry & theory of operation — the half-wave dipole

A half-wave dipole is two λ/4 conductors fed at the center, with a feedline routed perpendicular to the wire for at least a quarter wavelength to keep the coax shield from radiating. The geometry is one of the few in antenna practice where the physics, the construction, and the cost are simultaneously simple — which is exactly why this is the antenna you should build first and reach for most often.

2.1 Standing waves on a half-wave element

A half-wave conductor fed at its center is a resonator. At resonance the current distribution is a half-cycle sinusoid (maximum at the feedpoint, zero at the ends) and the voltage distribution is its 90°-shifted partner (zero at the feedpoint, maximum at the ends). The radiation comes from the accelerated charge — peak current is what radiates, so the center of a dipole is where most of the radiation originates, not the ends. This is the reason the antenna’s near-field pattern, common-mode behavior, and best feeding strategy are all centered on the geometric center.

      voltage peak                                   voltage peak
              \                                     /
               \                                   /
       END  ────●═══════════════●═══════════════●──── END
       (high V)        ↑                       (high V)
                     FEEDPOINT
                  (high I, low V)
                  ↑
            73 + j42 Ω free space

Two things follow directly from this picture. First: the BALUN belongs at the feedpoint, because that is where the coax shield’s outer surface ties into the antenna’s RF current path and where common-mode currents either propagate or get killed. Second: the ends want voltage and the center wants current, so the ends need insulation (dielectric strength) and the center needs conductivity (low-loss connection to the feedline) — these aren’t decorative choices, they’re the dominant failure modes a homemade dipole has to address.

2.2 The end-effect trim factor

A free-space half-wavelength at 7.15 MHz is 20.97 m. A wire dipole cut for 7.15 MHz is not 20.97 m; it’s about 19.92 m. The difference is the end-effect: capacitive coupling between the wire ends and surrounding insulators (and to a smaller extent the surrounding air and ground) lowers the velocity of propagation at the ends and shortens the resonant length. The trim factor k collapses this into one number:

L_physical = k · λ/2

where k is a function of the wire’s length-to-diameter ratio:

Wire dimension	length/diameter	k	Notes
#18 magnet wire	25,000+	0.97	Very thin wire, minimal end-effect
#14 / #12 antenna wire	5,000–15,000	0.95	The standard HF dipole — `468 / f_MHz` (feet) formula assumes this
#10 stranded	3,000–6,000	0.94	Slight fattening trim
3/16″ tubing	800–1,500	0.92	Fat element — used on 6 m and up
1/2″ tubing	300–800	0.89	Very fat — VHF dipoles, low-Q wide-bandwidth designs

The canonical formula for a thin-wire HF dipole is L_feet = 468 / f_MHz (which is k = 0.95 baked in), and it’s deliberately conservative — you cut a few percent long, sweep with the analyzer, and trim from both ends equally to bring the resonance down to where you want it. Always cut long. A pair of pliers makes wire shorter; nothing makes it longer except splicing, and a splice in the middle of a radiating element is a half-wave headache.

2.3 Half-wave length table — every band of interest

Cut these lengths, then trim with the analyzer. The two columns are the standard 95%-k formula (good for #14 wire) and the “cut long for trimming” length that gives you ~1% headroom.

Band	Design freq	λ/2 free-space	Cut at k=0.95	Cut long (+1%)
160 m	1.900 MHz	78.94 m (259.0 ft)	74.99 m (246.1 ft)	75.74 m (248.5 ft)
80 m	3.650 MHz	41.09 m (134.8 ft)	39.04 m (128.1 ft)	39.43 m (129.4 ft)
60 m	5.358 MHz	27.98 m (91.81 ft)	26.58 m (87.21 ft)	26.85 m (88.08 ft)
40 m	7.150 MHz	20.97 m (68.81 ft)	19.92 m (65.36 ft)	20.12 m (66.01 ft)
30 m	10.125 MHz	14.81 m (48.59 ft)	14.07 m (46.16 ft)	14.21 m (46.62 ft)
20 m	14.175 MHz	10.58 m (34.70 ft)	10.05 m (32.97 ft)	10.15 m (33.30 ft)
17 m	18.118 MHz	8.275 m (27.15 ft)	7.861 m (25.79 ft)	7.940 m (26.04 ft)
15 m	21.225 MHz	7.064 m (23.18 ft)	6.711 m (22.02 ft)	6.778 m (22.24 ft)
12 m	24.940 MHz	6.011 m (19.72 ft)	5.710 m (18.74 ft)	5.767 m (18.92 ft)
10 m	28.500 MHz	5.260 m (17.26 ft)	4.997 m (16.39 ft)	5.047 m (16.56 ft)
6 m	50.250 MHz	2.984 m (9.79 ft)	2.834 m (9.30 ft)	2.863 m (9.39 ft)
2 m	146.00 MHz	1.027 m (3.37 ft)	0.975 m (3.20 ft)	0.985 m (3.23 ft)
70 cm	446.00 MHz	0.336 m (1.10 ft)	0.319 m (1.05 ft)	0.323 m (1.06 ft)
33 cm	915.00 MHz	0.164 m (6.45 in)	0.156 m (6.13 in)	0.157 m (6.19 in)
23 cm	1295.0 MHz	0.116 m (4.56 in)	0.110 m (4.33 in)	0.111 m (4.38 in)

Tip. For VHF and up, build a dipole out of brass rod or copper tubing rather than wire — the trim factor stabilizes (fat-element behavior at high frequencies makes the percent-trim less sensitive to surroundings), and the wire’s mechanical Q at 50 MHz on up makes it droop in ways that shift the resonance after you’ve tuned it.

2.4 Why a dipole radiates — current sheet, not “wave bouncing”

A common mental model is “the RF wave bounces back and forth at the ends” — which is mathematically equivalent but obscures what’s happening physically. The truth is simpler: an accelerating charge radiates, and the standing-wave current distribution on a half-wave element accelerates the most charge in the most-uniform-phase configuration of any simple antenna geometry. The resulting far-field pattern (the figure-8 broadside to the wire that we’ll detail in §4) falls out of integrating the current distribution against the geometric phase delay across the antenna’s length.

The practical consequence: any modification that flattens the current distribution — folded dipole, fat element, coil-loaded compact dipole — broadens the bandwidth (less stored reactive energy per unit radiation) at modest cost to peak gain. Any modification that distorts the current distribution toward the ends — a top-hat-loaded vertical, an inverted-L — loses gain. The half-wave dipole’s clean current distribution is why it sets the reference.

3. Feedpoint impedance and matching

3.1 Free-space impedance: 73 + j42.5 Ω

A half-wave dipole in free space, cut to exactly a free-space half-wavelength, presents 73 + j42.5 Ω at the feedpoint. The 73 Ω real part is the radiation resistance (plus a tiny ohmic resistance from the wire). The +j42.5 Ω inductive reactance is leftover from the wire being slightly long for true resonance — the end-effect from §2.2 isn’t yet trimmed out.

Two ways to deal with the +j42.5: trim the wire shorter to bring the resonant length down (the standard approach — get the trim factor right, the reactance disappears with the length cut), or feed the longer wire and tune the reactance out with a series capacitor or a tuner. Almost no one does the second; the first is mechanical and stays put.

After trimming to resonance the feedpoint sits at 73 + j0 Ω in free space.

3.2 Real-ground impedance: 50–70 Ω

A dipole in free space is a textbook abstraction. Real dipoles sit above real ground at finite height, and ground reflection shifts the feedpoint impedance through induced currents on the wire. The shift is height-dependent:

Height above ground (h/λ)	Real Z (Ω)	Reactance (Ω)	Notes
0.10 (very low)	25–35	small +j	Strong ground coupling; lossy if not over excellent ground
0.20	45–55	small +j	Common NVIS height; works fine on 50 Ω coax
0.25	55–65	~0	Sweet spot — natural near-50 Ω match
0.35	65–72	~0	Approaching free-space
0.50	70–78	~0	The high-but-still-grounded case; 1.5:1 SWR typical
1.00	73	~0	Effectively free space
2.00+	73	~0	Free space

A 40 m dipole at 33 ft (10 m) is at h = 0.24λ — sweet spot, ~60 Ω, near-perfect 50 Ω match. A 40 m dipole at 60 ft is at 0.44λ, ~70 Ω, SWR 1.4:1. A 20 m dipole at the same 33 ft is at h = 0.47λ, ~70 Ω, also SWR 1.4:1. The pattern that emerges is that most practical horizontal-dipole heights land between 40 Ω and 75 Ω feedpoint resistance, which is why we tolerate the modest 50 Ω coax mismatch and don’t bother with a transformer for HF dipoles.

3.3 The 50 Ω coax mismatch — and why we don’t care

SWR for a 73 Ω resistive load on 50 Ω coax is 73/50 = 1.46:1. The mismatch loss is:

ML (dB) = -10 · log₁₀(1 - |Γ|²)

where |Γ| = (Z - 50) / (Z + 50) = 0.187, giving ML = 0.155 dB — about 1.5% power lost to the mismatch (the rest reflects into the rig where the source impedance absorbs it). This is a number nobody can hear on the air. The 1.46:1 SWR is also well within the protection range of every modern HF rig.

The instinct to “tune for 1.0:1” comes from QRP tuner-obsession and from a historical era when transmitters had narrow output matching ranges. With a modern solid-state HF radio, the rig will deliver full power into 1.5:1 without folding back; there is no operational reason to lose sleep over the 1.46:1 a clean 73 Ω dipole presents on 50 Ω coax.

3.4 The 1:1 current BALUN — at the feedpoint, always

A coaxial cable has three currents: the inner-conductor current, the inner surface of the shield’s current (these two are equal and opposite, by skin effect), and the outer surface of the shield’s current. The first two are the desired transmission-line currents. The third is unwanted: it’s a common-mode current that flows on the shield exterior, makes the shield into part of the antenna, distorts the radiation pattern, couples RF into the shack, and causes RF-in-the-shack symptoms (RF burns on the mic, distorted audio, intermittent ground-fault interrupters).

The 1:1 current BALUN (Guanella topology) at the feedpoint kills this common-mode current by presenting a high impedance to it while remaining transparent to the desired differential-mode currents. The mechanism is a transmission-line transformer wound on a ferrite core — the differential-mode current passes through unimpeded, but the common-mode current has to fight the choking inductance of the wound assembly.

BALUN type	Topology	When to use	Common-mode rejection
1:1 current (Guanella)	Transmission-line transformer, bifilar-wound	Centered-fed dipole, every half-wave	25–35 dB at HF
Voltage BALUN (Ruthroff)	Auto-transformer wound for voltage ratio	Older designs; not preferred	<15 dB CMRR
String-of-beads (W2DU)	50–80 Mix-43 ferrite beads slid over coax	Field-portable, replaces a discrete BALUN	20–30 dB
Air-core common-mode choke	6–10 turns of coax through 4–6″ form	Brute-force CMC; works without ferrite	15–25 dB
4:1 / 6:1 / 9:1 BALUN	Voltage or current with step-down	Multi-band dipoles (Vol 7) and EFHW (Vol 10)	varies

For a single-band half-wave dipole the answer is always 1:1 current BALUN at the feedpoint, on a 2.4″ Mix-31 (or 2.4″ Mix-43 for higher VHF reach) toroid with 12–15 bifilar turns of #14 enamel wire, or one of the commercial parts:

Balun Designs 1115a (1.5 kW SSB, Mix 31, $80, the canonical reference part)
DX Engineering DXE-BAL050-H10-A (1 kW, Mix 31, $75)
MFJ-918 (1.5 kW, $50, lower build quality — usable but tighter spec)
Palomar Engineers PAR-EF-MK2 (HF 1.8–30 MHz, $90)

A homebrew Mix 31 1:1 BALUN with 12 turns of #14 enamel on a 2.4″ FT240-31 toroid runs about $25 in parts (the toroid is the main cost — Amidon, Palomar, or eBay surplus) and handles 1.5 kW with no fan. Full winding instructions and the ferrite-mix selection criteria are covered in Vol 16 §3.

3.5 The matching summary

A trimmed-to-resonance half-wave dipole at residential height needs nothing more than a 1:1 current BALUN at the feedpoint. No tuner, no matching network, no series cap, no shunt coil. Build it, hoist it, sweep it, trim it, hoist it back, log the SWR, transmit.

The only nuance worth flagging: if your dipole sits at h < 0.15λ (very low — typical for stealth antennas tucked into an attic or under a roof eave), the feedpoint Z can drop into the 20–35 Ω range and the SWR climbs to 2:1 or worse. At that point a 1.5:1 unun (Balun Designs 1112a) at the feedpoint cleans it up.

4. Radiation pattern — azimuth, elevation, polarization

4.1 Free-space pattern — the donut

In free space (no ground, no nearby objects) a half-wave dipole has a torus pattern: broadside maximum, deep null off each end (>30 dB), peak gain 2.15 dBi. The 3D shape is a doughnut with the wire running through the hole; in azimuth (looking down on a horizontal dipole) the pattern is the familiar figure-8; in elevation (looking from the side) it’s a circle.

   Free-space pattern, horizontal dipole, looking down (azimuth):

                       N
                       ↑
                  ╱─────────╲
                ╱             ╲
              ╱      MAX        ╲
            ●   broadside (N-S)  ●
              ╲                 ╱
                ╲             ╱
                  ╲─────────╱
              W ───┤ wire ├─── E
                  ╱─────────╲
                ╱             ╲
              ●   broadside     ●
              ╲                 ╱
                ╲             ╱
                  ╲─────────╱
                       ↓
                       S

       deep nulls off the ends (W and E)
       maximum broadside (N and S)

4.2 Over real ground — the lobing pattern

A horizontal dipole over real ground does not radiate as if in free space. Ground reflection adds (or subtracts) coherently from the direct radiation, and the elevation pattern decomposes into a series of lobes. The peak elevation angle is what you actually care about — that’s where the propagation modes you want to reach are.

Height above ground (h/λ)	Peak elevation angle	Peak gain over ground	Propagation mode favored
0.05	90° (zenith)	5.5 dBi	NVIS — local/regional, 0–500 km
0.10	85°	6.5 dBi	NVIS, very strong
0.125	75°	7.0 dBi	NVIS / regional
0.20	60°	7.4 dBi	NVIS-to-regional transition
0.25	50°	7.5 dBi	Regional / short-skip
0.30	42°	7.6 dBi	Regional + opening DX path
0.40	32°	7.7 dBi	Mid-distance, useful for sky-wave DX
0.50	28°	7.8 dBi	DX-capable; first secondary lobe @ 90°
0.75	18°	7.5 dBi	Low-angle DX; secondary lobe @ 60°
1.00	14°	7.4 dBi	Low-angle DX; multiple useful lobes
1.50	9°	7.0 dBi	Very low angle, contest DX heights
2.00	7°	7.0 dBi	Approaching free-space pattern shape

For an HF dipole on a typical residential lot at 35–50 ft (about 0.25–0.35λ on 40 m, 0.5–0.75λ on 20 m), the pattern is “useful for everything from NVIS to medium-DX” — which is why a 40 m dipole at 40 ft is the universal first HF antenna. It does too many things at once to be optimal at any one of them, which is exactly the right behavior for a beginner antenna.

4.3 NVIS — the low-dipole DX-of-locals advantage

Near-Vertical Incidence Sky-wave is a niche-but-valuable propagation mode for 40 m and below: launch the signal nearly vertically, the ionosphere reflects it back down within a ~500 km radius, and the signal lands with a uniform footprint that fills in the “skip zone” between ground-wave range (~50 km) and short-skip (~700 km). A horizontal dipole at h = 0.1–0.15λ is the optimal NVIS antenna — its near-zenith maximum is exactly what NVIS wants.

For emergency communications, regional nets, and “talking to your friend two states over on 40 m when 20 m is closed,” a 40 m dipole at 15 ft beats a 40 m dipole at 50 ft every time. The high antenna nulls out NVIS angles; the low antenna is NVIS. This is the rare case where lower is better, and it’s worth knowing because it pushes back against the universal “higher is better” instinct.

4.4 Azimuth pattern — figure-8, with usable nulls

Looking down on a horizontal dipole the azimuth pattern is figure-8 with peak broadside and ~20–30 dB deep nulls off the ends. The nulls are clean enough to be useful for direction-finding (rotate the dipole until the signal disappears — that direction is along the dipole’s axis ± 180°) and they’re sharp enough that orienting a dipole carelessly can put a desired DX path in the null. For a north-south wire, peak gain is east-west; for an east-west wire, peak gain is north-south.

The nulls are deeper in free space than they are over ground. Real-ground operation softens the nulls to about 15–18 dB, still useful but not the textbook 25+ dB. A dipole over poor ground (sandy, dry, rocky) has worse null-depth than the same dipole over salt-water-class conductivity.

4.5 Polarization — and why you can’t easily use both

A horizontal dipole radiates horizontally polarized RF. A vertical dipole radiates vertical. Cross-polarization loss (the receiver expects horizontal, you’re transmitting vertical) is typically 18–20 dB in free space and 6–12 dB over real ground (where multipath partially scrambles the polarization). The cross-link to Vol 2 §5 covers the polarization-modal trade-offs in depth; for this volume the key facts are:

HF below 30 MHz: horizontal is the standard for fixed stations; vertical is the standard for mobile and for very-low-height (< 0.15λ) NVIS where the horizontal dipole’s pattern is unfavorable.
VHF/UHF FM: vertical is the standard (repeaters and handhelds are vertically polarized; cross-pol loss eats your link if you use horizontal).
VHF/UHF weak-signal (SSB, CW, FT8, EME): horizontal is the standard (the 6 m and 2 m weak-signal community uses horizontal Yagis universally).

A dipole’s polarization is dictated by the wire’s orientation. A horizontal dipole is horizontally polarized; tipping it 45° gives slant polarization (a compromise that’s mostly useful when working both horizontal and vertical contacts and not wanting to optimize for either).

5. Frequency response & SWR curve

5.1 The bandwidth story

A half-wave dipole is a Q ≈ 50 series-resonant LC circuit with the radiation resistance as the loss. Bandwidth (the 2:1 SWR points) is roughly:

BW (Hz) ≈ fc / Q

For a thin-wire (#14) HF dipole, Q ≈ 40–60 and the 2:1-SWR bandwidth is about 3–5% of the design frequency. Translated to band coverage:

Band	Design freq	2:1-SWR BW (#14 wire)	Coverage
160 m	1.900 MHz	50–80 kHz	<1/2 of CW + part of phone
80 m	3.650 MHz	130–180 kHz	Most of CW or most of phone — not both
40 m	7.150 MHz	250–350 kHz	Full band (well within 7.000–7.300)
20 m	14.175 MHz	600–800 kHz	Well over the 350 kHz band
10 m	28.500 MHz	1.5–2.0 MHz	Full band easily
6 m	50.25 MHz	2.5–3.5 MHz	Full band easily

The trap that bites everyone: on 80 m a single-band dipole only covers about a third of the band at SWR < 2:1. Cut for 3.650 MHz, the antenna sits inside the CW segment (3.500–3.600) at 2.5:1 and the phone-band edge (3.900) at 3:1 — usable with a tuner, painful without. For 80 m, the standard fixes are:

A dipole cut for CW (3.550) and accepting a tuner on phone, or vice versa.
Two separate dipoles (one CW, one phone) on the same feedpoint — see Vol 7’s fan dipole.
A wider-bandwidth element: cage dipole (Vol 7 §9), or folded dipole (§6.1 below).
Resignation and a tuner.

5.2 SWR curve shape

The SWR vs frequency curve for a thin-wire dipole is V-shaped with a sharp minimum at the resonant frequency. A typical 40 m dipole cut for 7.15 MHz shows:

SWR vs frequency, 40m dipole, #14 wire, h = 0.25λ, fc = 7.15 MHz:

  SWR
   3 ┤●                                       ●
     │  ●                                  ●
   2 ┤    ●                              ●
     │      ●                          ●
     │        ●                      ●
   1.5 ┤        ●                  ●
       │          ●              ●
   1   ●━━━━━━━━━━━━━●▼●━━━━━━━━━━━━━━━━━━●  ← min @ 7.15
       └─────┼──────┼──────┼──────┼──────┼─→ MHz
            6.8    7.0    7.15   7.3   7.5
                   |◄─ 2:1 bandwidth, ~330 kHz ─►|

For a fat-element dipole (3/16″ tubing) the curve flattens — the same antenna becomes 2:1 over ~500 kHz with a less-sharp minimum. For a folded dipole (§6.1) it widens further to 800 kHz–1 MHz. For a cage dipole (Vol 7 §9) it widens to 1.5–2 MHz with the same trade-off (more mechanical complexity, more wind load, more cost).

5.3 The trim sensitivity

Quantitatively, a 40 m thin-wire dipole shifts about 1 kHz down in resonance per 1.5 cm of wire added per side. Equivalently: trimming 1.5 cm off each end shifts resonance up by 1 kHz. This number scales linearly with frequency — at 14 MHz it’s about 0.7 cm per kHz, at 28 MHz it’s about 0.4 cm per kHz, at 144 MHz it’s about 1 mm per 100 kHz. The take-away is trim small — the temptation to “just chop off 6 inches and re-sweep” almost always overshoots, and re-splicing wire to lengthen the dipole is the kind of frustration you should not have to suffer in the field.

The trim procedure with a NanoVNA:

1. Cal the NanoVNA: SOLT calibration with the SMA fixture.
2. Connect the BALUN feedpoint via a short coax pigtail to port 1.
3. Sweep ±15% of design frequency (e.g. 6.0–8.3 MHz for a 40 m design).
4. Read the resonant frequency f_res (the SWR minimum, with R near pure resistance on the Smith chart).
5. Δf = f_design - f_res
6. Calculate the trim:  trim_per_side ≈ Δf · (cm/kHz factor from the table below)
7. Trim equal amounts from each end; re-tension; re-sweep.
8. Iterate until f_res lands within ~10 kHz of f_design.

Per-band trim sensitivity for #14 wire:

Band	cm trim per side per kHz of upward shift
80 m	5.8 cm/kHz
40 m	1.5 cm/kHz
30 m	0.75 cm/kHz
20 m	0.40 cm/kHz
15 m	0.18 cm/kHz
10 m	0.10 cm/kHz
6 m	0.03 cm/kHz
2 m	0.003 cm/kHz (need a file, not a wire cutter)

Tip. Always cut and trim both sides equally. Asymmetric trim shifts the current distribution off the geometric center and degrades the figure-8 pattern by 1–3 dB; the SWR will still bottom out, but the pattern won’t be clean.

6. Variants — folded, inverted-V, sloper, vertical dipole

These four variants share the half-wave-dipole core but bend the geometry, feedpoint, or polarization in useful ways. Each is still single-band; the multi-band relatives are in Vol 7.

6.1 Folded dipole

A folded dipole is two parallel half-wave wires shorted at the ends, fed across the gap in one of them at the center. The feedpoint impedance jumps to ~280 Ω (free space) — exactly 4× the standard dipole’s 73 Ω because the same current flows in both parallel conductors and the radiation pattern is unchanged.

  feedpoint ●═══════════════╤═══════════════● (shorted ends)
            ║              ╧                ║
            ●━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━●
            ⇕ ~50 mm spacing (or 300 Ω twin-lead)

What you get:

2–3× bandwidth vs straight dipole (Q drops from ~50 to ~20)
Slightly higher gain (~0.1 dB — negligible)
Robust matching via 4:1 BALUN to 50 Ω coax (or direct 300 Ω feedline to a balanced tuner)
The mechanical complication of two parallel conductors at constant spacing — usually built from 300 Ω twin-lead or 450 Ω ladder line with the parallel conductors shorted at the ends, or from two #14 wires with spreaders every 30 cm.

Where the folded dipole wins: 80 m, where the extra bandwidth lets you cover CW + phone without retuning, and FM-broadcast / VHF receive where the wider bandwidth covers an entire band segment with one element.

6.2 Inverted-V

An inverted-V is a half-wave dipole with both ends drooped from a single central support — the apex is high, the ends are low. The apex angle is typically 90–120° (each leg drooping 45–60° from horizontal). What changes:

Feedpoint Z drops from 73 Ω to about 50–60 Ω depending on apex angle — natural 50 Ω match with no transformer needed
Pattern broadens — the figure-8 nulls fill in, gain drops 0.5–1 dB, but the antenna becomes more omnidirectional which is often what you want
Only one support needed — single mast, single rope, single tree
Footprint shrinks — the ends can be staked closer to the support point than the horizontal extent demands for a flat dipole

The inverted-V is, frankly, what most amateurs actually build when they say “I put up a dipole.” The single-mast convenience drives the choice; the pattern-broadening and the small gain loss are acceptable for general HF operation.

Apex angle and its consequences:

Apex angle	Each leg droop	Z at feedpoint	Pattern	Notes
180° (flat dipole)	0°	73 Ω	Sharp F8 nulls, 7.8 dBi peak	Reference
120° (mild V)	30°	65 Ω	Soft F8, 7.5 dBi peak	Best compromise
90° (standard V)	45°	50 Ω	Soft pattern, 7.0 dBi peak	The conventional inverted-V
60° (steep V)	60°	40 Ω	Nearly omni, 6.5 dBi peak	Only when mast height forces it
30° (extreme V)	75°	25–30 Ω	Omni, 5.5 dBi peak	Approaches vertical dipole behavior

6.3 Sloper

A sloper is a dipole tilted off horizontal — one end elevated, the other end low. The most common configuration is “feed-point at the top of a tower, far end going down at 30–45° to ground.” What changes:

Pattern skews toward the lower end — gain is enhanced in the direction of the low end (~2–3 dB lift) and reduced in the opposite direction
Vertical polarization component appears — the slope means part of the antenna’s current is vertical, so the radiation includes a vertical-pol component (favored for DX)
Feedpoint Z drifts depending on slope angle — typically 40–60 Ω, often 50 Ω natural match
Mast/tower interaction — sloping off a tower means the tower’s own RF currents matter; ground the tower well and current-choke the coax aggressively

The sloper is a good directional fixed-installation antenna when you want to skew radiation toward a specific direction (a favored DX path, or away from a noisy neighbor) and you have a single tall support. Less popular today since proper directional Yagis are affordable; still useful as a low-cost direction-favoring single-band antenna.

6.4 Vertical dipole

A vertical dipole is a half-wave dipole oriented vertically — either a continuous wire/tube fed at the center, or an end-fed half-wave with a sleeve choke that effectively makes the lower λ/4 of the feedline radiate as the second half. Differences from the horizontal dipole:

Vertical polarization — useful where vertical pol is dominant (mobile/handheld VHF/UHF, DX paths on HF where ground reflection is less of a friend)
Omnidirectional azimuth — no figure-8 nulls; this is the dipole pattern’s “polar opposite” — instead of broadside max with end nulls, you get an azimuth-uniform donut with a vertical null straight up
No radial field required — unlike a quarter-wave monopole (Vol 8), a half-wave dipole’s currents are self-contained; the ends sit at high impedance and don’t need an image plane to function. This is the vertical dipole’s killer feature for portable VHF/UHF operation.
Feedpoint Z the same 73 Ω (free space) or 50–70 Ω over real ground
Mounting requires a non-conductive support running parallel to the wire — fiberglass or wood mast, not metal

The classic vertical-dipole implementations:

End-fed half-wave with a sleeve choke (W2DU configuration, sleeve = coax shield extended into a quarter-wave sleeve via a series choke at the feedpoint) — the way most commercial “no radial” verticals work at HF
Roll-up sleeve dipole for 2 m — Slim Jim and J-pole geometry (covered in Vol 9 §6) — a vertical dipole with a parallel quarter-wave matching stub
HF vertical dipole for 17/15/10 m where the half-wavelength is short enough that hoisting a vertical fiberglass mast with a wire dipole on it is practical (e.g. SteppIR vertical dipole; HyEndFed 17 m vertical-dipole)

7. Best-case use

A half-wave dipole earns its dominant-default-antenna status where:

Single-band resonant HF operation is the use case — 40 m, 30 m, 20 m, 17 m, 15 m all live happily in backyards. The bandwidth limitation on 80 m is the only place the dipole strains.
Field-portable single-band operation — the dipole rolls up small, deploys fast (toss a halyard over a branch, hoist, sweep, transmit), and works predictably with no ground system to deploy.
Receive direction-finding — the deep azimuth nulls (in free space ~30 dB, over real ground ~15–18 dB) make a horizontal dipole a useful HF DF antenna when paired with a rotator.
A first HF antenna — the half-wave dipole at 0.25–0.5λ height is the universal “first antenna” because it works well over a broad range of propagation modes (NVIS to medium-DX), needs no tuner if you stay close to resonance, costs $40 in parts, and teaches you the most by being predictable.
Reference for measurement — when characterizing a propagation path, or comparing antennas in an A/B test, a half-wave dipole is the standard. dBd as a unit only means anything if everyone agrees on what a dipole is.
Multi-element antenna building block — Yagi-Uda arrays (Vol 11), log-periodics (Vol 13), phased verticals — the basic radiating element of a half-wave (sometimes folded for impedance) is the half-wave dipole. Knowing how it behaves alone is prerequisite to building anything that uses it as a part.

8. Worst-case use

The half-wave dipole falls down where:

Multi-band operation without a tuner — single-band is the constraint. Two bands need two dipoles (or a fan dipole), three bands need three, and at four-plus bands the multi-band specials in Vol 7 start to win.
Stealth deployment — a 40 m dipole is 20 m of horizontal wire and visible from blocks away. HOA-restricted lots, apartment balconies, attic installations — see Vol 23 for the stealth playbook; the dipole is the wrong choice when you can’t show the antenna.
Low-angle DX at low residential heights — an HF dipole at h < 0.3λ has its peak elevation lobe above 40°. Working Europe from the US on 20 m at 35 ft means your peak gain is at 50° elevation — well above where the propagation path lives. The signal still gets through (the dipole has gain at 15° too, just not peak) but a vertical or a higher dipole would gain you 4–6 dB.
80 m and 160 m on residential lots — the size becomes a problem. A full-size 80 m dipole is 40 m of wire; full-size 160 m is 80 m. At those sizes the install is the limiting factor, not the antenna design — see Vol 7 for shorter multi-band options that compromise gracefully, or Vol 8 for vertical monopoles that fit smaller footprints.
High local noise environments — a horizontal dipole picks up less local-electrical noise than a vertical (vertically-polarized noise from power-line discharge, switched-mode supplies, plasma-TVs dominates the noise floor; horizontal polarization rejects it partly). But if your local noise is from a horizontal source (overhead PV inverters, neighbor’s power-line fault), the dipole receives it like any other antenna. There’s no “low-noise” antenna; only antennas that happen to be cross-polarized to your dominant local interference.

9. Power handling

A wire dipole’s power-handling limits sit in three places:

9.1 The wire

#14 AWG stranded copper carries:

Continuous power: ~1.5 kW SSB / 2 kW CW key-down without heating concerns
Pulse power: tens of kW (legal-limit amateur is well below the wire’s limit)
Mechanical: ~50 kg break strength — the wind-loading and ice-loading limits matter more than the RF current limits

Thinner wire (#18 magnet wire, #20 stealth wire) drops the limit to 100–200 W. Thicker wire (#12, #10) raises it to 3+ kW. For amateur HF, #14 is the sweet spot — 1.5 kW capable, costs $30 for 30 m, easy to terminate, mechanically robust.

9.2 The insulators

Ceramic end insulators (DX Engineering DXE-ISD, Glen Martin, MFJ-16C03) are good to ~30 kV — that’s far beyond what any legal-limit amateur dipole sees at the ends (at 1.5 kW into 73 Ω the standing-wave peak voltage at the ends is about 5 kV). Polyethylene-bodied insulators are good to ~20 kV — also far over the requirement. Don’t use kitchen-grade plastic; it tracks and arcs under RF voltage at far below its DC ratings.

9.3 The BALUN — usually the actual limit

A 1:1 current BALUN on a Mix-31 toroid (the workhorse choice) saturates at about 1.5 kW SSB / 1 kW continuous-key-down. Mix-43 saturates lower. Mix-61 handles VHF but is overkill for HF. Mix-52 (Fair-Rite’s HF / low-VHF mix) is gaining in popularity.

BALUN core	Mix	Continuous power rating
FT240-31 (2.4″ Mix 31)	31	1.5 kW SSB / 1 kW CW
FT240-43 (2.4″ Mix 43)	43	1 kW SSB / 800 W CW
FT240-52 (2.4″ Mix 52)	52	1.5 kW SSB / 1.2 kW CW
FT290-31 (2.9″ Mix 31, big)	31	3 kW SSB / 2 kW CW
String-of-beads (50× Mix 43)	43	1 kW SSB
Air-core CMC (10 turns on 4″ form)	n/a	5+ kW (limited by coax, not the choke)

For most amateur installations a 2.4″ Mix-31 toroid with 12 turns of #14 enamel is the right choice; for 1.5 kW + amplifier the 2.9″ toroid is the upgrade path. Full ferrite-mix selection theory is in Vol 16 §10.

10. DIY build — a 40 m half-wave dipole, step by step

This is the canonical “first antenna” build. About 90 minutes of work plus tuning time. Total parts cost ~$120 USD at mid-2026 list prices.

10.1 Bill of materials

Part	Specification	Source	Mid-2026 price
Antenna wire	#14 AWG copper, 7-strand, insulated or bare, ~24 m (78 ft)	DX Engineering DXE-ANTW-14B (insulated black PE jacket, $0.92/ft) or Wireman 534 (bare hard-drawn copper, $0.55/ft)	$22–35
End insulators	Ceramic egg-style, ~2″ long, 30 kV rated	DX Engineering DXE-ISD-25 (each $4.50) or Glen Martin EI-1 (each $5)	$9–10
Center insulator	Plastic body with SO-239 or N feedpoint, strain-relief eye	Budwig HQ-1 (SO-239, $18) or DX Engineering DXE-COA-1 (SO-239, $25)	$18–25
1:1 current BALUN	Mix 31 toroid, 1.5 kW SSB rated, weatherproof enclosure	Balun Designs 1115a ($80) or DIY (FT240-31 toroid + 12 turns #14 enamel, $25)	$25–80
Halyard rope	1/8″ Dacron/polyester, UV-stable, ~30 m total	Synthetic Textiles 1/8″ Dacron ($0.25/ft) or DX Engineering DXE-DACRON ($0.45/ft)	$15–30
Coax pigtail	RG-8X (or RG-58 for QRP), 1 m, SO-239 to PL-259 (or N)	Times Microwave RG-8X with field-installable PL-259 connectors	$15
Coax-Seal	Self-amalgamating tape, 1 roll	3M Scotch 130C (rubber tape) + 3M Super 33+ (vinyl overwrap)	$12
Total			$116–207

The cost spread is the BALUN: $80 commercial vs $25 homebrew. Both work; the commercial part saves an evening of toroid winding.

10.2 Step-by-step construction

Cut the wire. Total length for a 40 m dipole resonant at 7.150 MHz is 468 / 7.150 = 65.45 ft (~19.95 m). Cut two lengths of ~10.05 m each (~33 ft) — that’s about 0.5% long on each side, giving you ~50 kHz of trim headroom. Always cut long.

Terminate the ends. Each wire end loops through the ceramic insulator’s hole, doubles back ~15 cm, and is wrapped 6–8 turns around the standing wire then soldered. The doubled-back section is the strain-bearing loop; the soldering is electrical security against corrosion at the splice. A poorly terminated end will untwist in wind over months; a properly wrapped + soldered end is good for 20+ years.

Terminate the center. Each wire’s inboard end passes through a strain-relief eye on the center insulator, doubles back ~12 cm, and is soldered to the center-insulator terminal. The two wires must be electrically separate (they connect through the BALUN, not directly to each other). The center insulator’s terminals are spaced 5–8 cm apart, which is the dipole’s “gap” in §2.1’s standing-wave picture.

Install the BALUN. If using a commercial BALUN like the Balun Designs 1115a, the feedpoint terminals are bolts on a weatherproof enclosure; bond each wire to its terminal with a soldered lug. The BALUN’s SO-239 connector accepts the coax feedline. If using homebrew (12 turns #14 enamel on an FT240-31 toroid), the BALUN winding’s two pairs of leads are soldered: one pair across the antenna terminals on the center insulator (the “balanced” side), the other pair to the coax pigtail’s center conductor and shield (the “unbalanced” side). Pot the homebrew BALUN in a UV-stable enclosure (Hammond 1591-XX series in fiberglass or polycarbonate, $20).

Add halyards. Each ceramic end insulator gets a Dacron halyard of sufficient length to reach over a tree branch or up a mast and back to a tie-off cleat. Plan halyard length as 2 × (tree branch height + tie-off distance from tree) + 3 m service slack. Use a stop knot at the insulator end (figure-8 with a follow-through; bowline if you prefer).

Hoist and measure. Hoist the dipole to its operating height (40–50 ft / 12–15 m typical for 40 m). Connect the NanoVNA to the feedline at the rig end. Sweep 6.0–8.0 MHz. Read the SWR minimum and the frequency it lands on. If the minimum is below 7.150 (e.g. at 6.90 MHz), the wire is too long — trim. If above (e.g. at 7.30 MHz), the wire is too short — you cut too aggressively (this should not happen if you followed “cut long”).

Trim. Lower the dipole, snip an equal amount from each end (use the cm/kHz table in §5.3 — for 40 m, ~1.5 cm per side per kHz of upward shift you need). Re-terminate the ends. Re-hoist. Re-sweep. Iterate until the resonance lands within ~10 kHz of your target frequency. Two to three iterations is typical.

Lock and weatherproof. Once tuned, apply 3M Scotch 130C self-amalgamating tape to the BALUN-to-coax junction and any field-installed connectors. Overwrap with Super 33+ vinyl tape for UV protection. Apply Coax-Seal (sticky black mastic) around any connector that goes outdoors and won’t be serviced for a year. The weatherproofing protocol is covered in more depth in Vol 22.

10.3 Tuning verification

A successful build has these characteristics on the NanoVNA:

SWR minimum < 1.4:1 at the target frequency (better is rare and not meaningful)
Resistive impedance at minimum (Smith chart marker near the horizontal axis, X ≈ 0)
R ≈ 50–73 Ω at minimum, depending on height
2:1 SWR bandwidth matches the table in §5.1 (~330 kHz for #14 wire at 7.15 MHz)
No anomalous peaks between half-wavelength resonance and 3× the design frequency (an anomalous peak at, e.g., 7.5 MHz with the antenna cut for 7.15 indicates a coax common-mode resonance — the BALUN is doing its job poorly and the coax is part of the antenna; check the BALUN winding)

Warning. If your SWR minimum is at 14.3 MHz instead of 7.15 MHz, you cut a 20 m dipole. Re-measure. This is a common new-builder mistake (confusing 468/14.175 = 33 ft for the 40 m length) and the symptom is “antenna’s lowest SWR is at 2× the design frequency.”

11. Commercial buys — current good-quality manufactured options

Sorted by price tier; mid-2026 USD list prices. Vendor coverage is North-America-centric; European equivalents from W&L Tools, Wimo, and SOTABEAMS are noted where relevant.

Tier	Model	Bands	Price	Notes
Budget	MFJ-1640T	40 m mono	$35	Resonator dipole, OEM-fab build, modest BALUN. Adequate for casual use.
Budget	Hy-Power Antennas 40 m doublet	40 m	$45	Solid materials, basic construction.
Budget	SOTABEAMS Band Hopper	40/30/20 m linked	$90	EU-centric, link dipole, well-built.
Mid	DX Engineering Resonator Dipole	per-band	$120–180	Pre-cut, factory-tuned, ~3% bandwidth, marine-grade hardware.
Mid	Buckmaster 7300-OCF	8-band OCFD (covers Vol 7 territory)	$150	The reference OCFD for mid-tier; covered in Vol 7 §3.
Mid	W5GI Mystery Antenna	80–10 m multi-band	$130	Variant of a multi-trap dipole; works but compromises.
Mid	Diamond W-8010	80/40/20/15/10 m	$180	Pre-cut trap dipole. Reliable but lossy on traps (Vol 7 §5).
Premium	DX Engineering DX Flag Dipole Kit	per-band	$280–400	Premium build, marine hardware, custom-cut. The “I’ll buy quality once” option.
Premium	Force 12 monoband dipole	per-band	$300–450	Rotatable single-band dipoles; aluminum tubing. The DX contest-grade reference.
Premium	InnovAntennas single-band rotatable dipole	per-band	$300–500	UK manufacturer; clean rotatable build.
Premium	Cushcraft D-3 / D-4 / D-40	3-band / 4-band / 40m mono	$400–800	Trap-style multi-band dipoles; established product line.

What to avoid: “MFJ jumper-wire dipoles” re-spooled as new — the coax pigtail and the SO-239 connector are typically where corners are cut, and you’ll spend more time replacing the feedline connector than building from scratch. Unbranded “1 kW dipoles” on eBay/AliExpress that don’t quote a BALUN saturation rating — these usually have no BALUN at all, just a soldered-junction “direct feed” that puts common-mode current straight into your shack. Anything billed as “no soldering required” — the soldered terminations at the insulators and BALUN are exactly what gives the antenna its 20-year service life.

EU/UK alternatives (for completeness): W&L Tools (Germany), Wimo (Germany), Sotabeams (UK), Cobwebb (UK), G7FEK (UK) — all sell good-quality single-band and multi-band dipoles into the European market, with prices roughly equivalent to North American suppliers + shipping.

12. Companion gear

A dipole doesn’t operate in isolation. The supporting cast:

Mast or single support pole — for inverted-V deployment, a single fiberglass push-up mast (Spiderbeam 12 m / 18 m, Jackite 31′, MFJ-1916) or telescoping aluminum mast. Full mast catalogue in Vol 21.
Halyard rope — Dacron 1/8″ (Synthetic Textiles, DX Engineering DXE-DACRON, or USRopes). Not nylon — nylon stretches under load, and a stretched halyard means your antenna sags and detunes daily.
Feedline — RG-8X for short runs (<25 m) and 100 W operation; LMR-400 for longer runs or higher power; hardline (LDF4-50A) for permanent installations >50 m. Connector + loss budget in Vol 5 §6.
Center insulator strain relief — even with the BALUN’s strain-relief eye, a separate strain-bearing rope from the center insulator’s mounting point to the support takes mechanical load off the antenna wires and the BALUN itself.
Lightning protection — a polyphaser arrestor in the coax line, grounded at the shack entry, is mandatory for permanent installations. The arrestor model and the single-point ground topology are detailed in Vol 20 §5.

13. Common gotchas and myths

“A dipole has 2.15 dBi of gain” — yes, in free space. Over real ground at h = λ/2, the same dipole has ~7.8 dBi of gain at peak elevation (~28°) and 0 dBi at zenith. Most published “dipole gain” numbers either disagree (because they’re using different reference heights), or they’re free-space figures being misapplied to over-ground antennas. The right answer is “it depends on h/λ” — read it off §4.2’s table for your installation height.
“I’ll just throw it up between two trees” — trees move in wind, sway under ice, and bend under snow. A wire between two trees with no strain-relief absorbs every meter of tree motion into wire tension, and after a winter you’ll find broken wire at the insulators. Use a halyard system: rope through a pulley over the branch, antenna wire bowline-tied to the rope, counterweight at the ground end (a 2 kg sandbag) that lifts to absorb tree motion. The antenna stays at constant tension; the tree absorbs the motion through the rope.
“The MFJ 5-band dipole works on 5 bands” — only if you don’t mind a tuner and 6 dB of off-resonance loss on four of them. Multi-band dipoles with traps or links work, but they compromise. The honest multi-band approach (fan dipole, doublet with tuner, OCFD) is in Vol 7.
“50 Ω coax is wrong for a 73 Ω dipole” — the 1.46:1 SWR penalty is 0.13 dB. Nobody can hear 0.13 dB on the air. Stop worrying about “matching for performance”; worry about matching for power-transfer-into-the-rig (which 1.5:1 satisfies fine for any modern HF rig).
“My SWR is 1.0:1 at the rig” — if you measure SWR at the rig and it’s exactly 1.0:1, your feedline loss is hiding mismatches at the antenna end. A noisy 6 dB feedline loss turns any antenna load into a near-50 Ω presentation at the rig — including a dummy load. Always measure SWR at the feedpoint (or at least at the BALUN side of the coax) to know what the antenna actually looks like.
“Resonance = best SWR” — not always. Resonance means X = 0 (pure resistance). Best SWR means the resistance is closest to 50 Ω. A high-Z resonant point (e.g. 200 Ω at resonance) still shows 4:1 SWR. The two coincide for a dipole near its design frequency, but the distinction matters for end-fed half-waves and other multi-resonance antennas where the lowest-SWR point isn’t always at resonance.
“Common-mode current is a theoretical concern” — no, it’s a real symptom: garbled audio at the receive end, RF burns when you key up while touching anything metallic in the shack, mysterious “the SWR is low but the antenna doesn’t seem to be playing” episodes. A 1:1 current BALUN at the feedpoint kills it; “I’ll add one later” is a year of degraded operation.
“Higher is always better” — for medium-DX (1500–6000 km), yes. For NVIS (regional 0–500 km on 40 m and below), no — lower is better. A 40 m dipole at 15 ft is the NVIS-optimal height; the same antenna at 50 ft has a null where you want maximum gain. The “always higher” instinct is the right default but not the right answer for every use case.

14. Resources

ARRL Antenna Book Ch. 6 (dipole and inverted-V) — the canonical 25th-edition reference, ~$50, every antenna-builder’s first investment.
Balanis, Antenna Theory (4th ed.) — the academic reference for radiation pattern derivations.
DX Engineering Tech Articles — dipole construction notes (free, online).
L. B. Cebik (W4RNL) papers on dipole modeling — archived at the antenneX library; the deepest treatment of dipole-over-ground behavior in the amateur literature.
Sevick, Transmission Line Transformers (5th ed.) — for the BALUN designer who wants to roll their own.
NanoVNA-Saver (PC software for the NanoVNA) — https://github.com/NanoVNA-Saver/nanovna-saver; the antenna-builder’s free workstation.

Contents