Skip to content
ELECTRONICS ENGINEERINGelectronics-engineering.com
Antennas Electromagnetic Fields

Radiation resistance: elementary dipole, short dipole, the 73 Ω of the half-wave dipole proved, monopole and efficiency

Radiation resistance step by step: elementary dipole, short dipole with triangular current, half-wave dipole with the cosine integral Cin(2π), directivity 1.64, monopole, dipole of any length, skin-effect losses and efficiency of short antennas, in Python.

https://www.electronics-engineering.com/en/radiation-resistance-dipole-73-ohm-proof-monopole-efficiency/

Was this page useful? Rate it:
average 4.0 from 1 rating

The radiation resistance R_r is the number that tells how well an antenna “gets rid” of power in the form of waves: it is the resistance that, carrying the antenna current, would dissipate the radiated power, P = \frac{1}{2}R_rI_0^2. It decides the match to the cable (73 Ω for the half-wave dipole, 36.6 Ω for the monopole) and the efficiency: a short antenna has a tiny R_r and almost all the power is lost as heat. This in-depth study derives R_r for all the basic wire antennas, with full proofs: the elementary dipole, the short dipole with triangular current, the half-wave dipole with the exact calculation of 73.1 Ω through the cosine integral, the monopole, directivity and the dipole of any length; then ohmic losses and efficiency. In the original version of this page the calculation was done with Mathematica; here it is in Python, which is free, with verified code.

The outline of the calculation

For every antenna the procedure is the same:

1. fix the current distribution along the wire;

2. compute the far field: each segment dz is an elementary dipole and the contributions add with their phase;

3. integrate the power density \frac{|E|^2}{2\eta} over a sphere, obtaining P;

4. set R_r = \frac{2P}{I_0^2}, where I_0 is the reference current (at the maximum or at the terminals: for the half-wave dipole they coincide).

Here \eta = \sqrt{\mu_0/\varepsilon_0} \approx 120\pi \approx 377 Ω is the impedance of free space and k = 2\pi/\lambda.

The three current distributions
On the left the elementary dipole, with uniform current over a length dl; in the centre the short dipole, whose current decreases linearly from the centre to the ends; on the right the half-wave dipole, with sinusoidal current. The small circle is the feed point.

The elementary dipole

A segment dl \ll \lambda with uniform current I_0 radiates in the far field

    \[ E_\theta = j\eta\,\frac{kI_0\,dl}{4\pi r}\,\sin\theta\,e^{-jkr} \]

(derived in the radiation lesson from the vector potential). The power is

    \[ P = \int_0^{2\pi}\!\!\int_0^\pi \frac{|E_\theta|^2}{2\eta}\,r^2\sin\theta\,d\theta\,d\varphi = \frac{\eta k^2I_0^2\,dl^2}{32\pi^2}\cdot 2\pi\int_0^\pi\sin^3\theta\,d\theta \]

With \int_0^\pi\sin^3\theta\,d\theta = \frac{4}{3} (write \sin^3\theta = \sin\theta(1 - \cos^2\theta) and set u = \cos\theta): P = \frac{\eta k^2I_0^2dl^2}{12\pi}. Hence

    \[ R_r = \frac{2P}{I_0^2} = \frac{\eta k^2dl^2}{6\pi} = \frac{2\pi\eta}{3}\left(\frac{dl}{\lambda}\right)^2 = 80\pi^2\left(\frac{dl}{\lambda}\right)^2\ \Omega \]

The short dipole

A real short dipole (of length l \ll \lambda) does not have uniform current: at the ends of the wire the current must vanish, and on a short wire it decreases almost linearly, I(z) = I_0\left(1 - \frac{2|z|}{l}\right). Since all segments are close compared with \lambda, the phases of the contributions are equal and the field is that of an elementary dipole with moment \int I(z)\,dz = I_0\frac{l}{2}: the effective length is half the physical one. Substituting dl = l/2:

    \[ R_r = 80\pi^2\left(\frac{l}{2\lambda}\right)^2 = 20\pi^2\left(\frac{l}{\lambda}\right)^2\ \Omega \]

four times less than the elementary dipole of the same length. A \lambda/10 dipole has about 2 Ω.

The half-wave dipole: the 73 Ω proved

With current I(z) = I_0\cos kz for |z| \le \lambda/4, adding the elementary dipoles with their phases e^{jkz\cos\theta} gives (dipoles lesson)

    \[ E_\theta = j\eta\,\frac{I_0}{2\pi r}\,\frac{\cos\left(\frac{\pi}{2}\cos\theta\right)}{\sin\theta}\,e^{-jkr} \]

The power is

    \[ P = \frac{\eta I_0^2}{4\pi}\int_0^\pi\frac{\cos^2\left(\frac{\pi}{2}\cos\theta\right)}{\sin\theta}\,d\theta \]

and the integral, often said to be “numerical only”, can be computed exactly.

Theorem. \displaystyle\int_0^\pi\frac{\cos^2\left(\frac{\pi}{2}\cos\theta\right)}{\sin\theta}\,d\theta = \frac{1}{2}\text{Cin}(2\pi), where \text{Cin}(x) = \displaystyle\int_0^x\frac{1 - \cos t}{t}\,dt is the cosine integral.

Proof. Set u = \cos\theta, du = -\sin\theta\,d\theta, and multiply and divide by \sin\theta: the integral becomes \int_{-1}^{1}\frac{\cos^2(\pi u/2)}{1 - u^2}\,du. With \cos^2\frac{\pi u}{2} = \frac{1 + \cos\pi u}{2} and the partial fractions \frac{1}{1 - u^2} = \frac{1}{2}\left(\frac{1}{1 - u} + \frac{1}{1 + u}\right):

    \[ \frac{1}{4}\int_{-1}^{1}(1 + \cos\pi u)\left(\frac{1}{1 - u} + \frac{1}{1 + u}\right)du = \frac{1}{2}\int_{-1}^{1}\frac{1 + \cos\pi u}{1 + u}\,du \]

because the change u \to -u turns the first term into the second. Now t = \pi(1 + u): \frac{du}{1 + u} = \frac{dt}{t}, \cos\pi u = \cos(t - \pi) = -\cos t, and the limits become 0 and 2\pi. The result is \frac{1}{2}\int_0^{2\pi}\frac{1 - \cos t}{t}\,dt = \frac{1}{2}\text{Cin}(2\pi). ∎

\text{Cin}(2\pi) = 2.4377 (the function is tabulated, and near t = 0 the integrand tends to zero, so there are no convergence problems). Hence

    \[ R_r = \frac{2P}{I_0^2} = \frac{\eta}{4\pi}\,\text{Cin}(2\pi) = 30\,\text{Cin}(2\pi) = 73.1\ \Omega \]

Note that the short-dipole approximation, 20\pi^2(0.5)^2 = 49 Ω, would be 33 % wrong here: at half a wavelength the phases of the contributions are no longer equal.

Directivity. The maximum radiation intensity (at \theta = 90°) is U_{max} = \frac{r^2|E_\theta|^2}{2\eta} = \frac{\eta I_0^2}{8\pi^2}, so

    \[ D = \frac{4\pi U_{max}}{P} = \frac{4\pi\,\eta I_0^2/(8\pi^2)}{R_rI_0^2/2} = \frac{\eta}{\pi R_r} = \frac{120}{73.1} = 1.64 \]

i.e. 2.15 dBi, the value used to define dBd. The same calculation for the elementary dipole, with U_{max} = \frac{\eta k^2I_0^2dl^2}{32\pi^2}, gives D = 1.5.

The monopole. A \lambda/4 monopole over a conducting plane produces, above the plane, the same field as the half-wave dipole (image theorem), but radiates only into the upper half-space: half the power with the same current, hence R_r = 36.6 Ω.

The dipole of any length

For a thin dipole of length l with current I_0\sin\big(k(l/2 - |z|)\big), the same procedure gives the resistance referred to the current maximum:

    \[ R_{max} = \frac{\eta}{2\pi}\int_0^\pi\frac{\left[\cos\left(\frac{kl}{2}\cos\theta\right) - \cos\frac{kl}{2}\right]^2}{\sin\theta}\,d\theta \]

The resistance seen at the terminals is R_{in} = R_{max}/\sin^2(kl/2), because for the same power the current at the centre is I_0\sin(kl/2). For l = \lambda the current at the centre vanishes and R_{in} becomes theoretically infinite: the centre-fed full-wave dipole is high impedance (in practice a few thousand ohms, because the wire has thickness).

Radiation resistance as a function of dipole length
The blue curve, referred to the current maximum, is 73.1 ohms at half a wavelength and 199 ohms at a full wavelength, with a maximum near 0.9 wavelengths; the dashed grey curve is the short-dipole approximation, valid only for small lengths.
l/\lambda0.10.250.511.5
R_{max} (Ω)0.196.7273.1199105
R_{in} (Ω)2.0013.473.1\infty105
20\pi^2(l/\lambda)^2 (Ω)1.9712.349.3

The short-dipole formula (referred to the terminal current) is good up to l \approx \lambda/10 and is already 8 % wrong at \lambda/4.

Losses and efficiency

The wire has an ohmic resistance R_p, and the supplied power splits between radiation and heat:

    \[ \eta_{rad} = \frac{R_r}{R_r + R_p} \]

Skin-effect losses. At radio frequencies the current flows in a surface layer; a wire of radius a and length l has resistance R = \frac{l}{2\pi a}R_s, with R_s = \sqrt{\pi f\mu_0/\sigma} the surface resistance. With sinusoidal current the dissipated power, referred to I_0, is half that with uniform current (the average value of \cos^2 is 1/2). For a copper half-wave dipole (\sigma = 5.8\cdot 10^7 S/m) at 100 MHz, with a = 1 mm: R_s = 2.61 mΩ, R_p = 0.31 Ω, efficiency 99.6 %.

The short antenna. A whip of h = 1 m over a ground plane at 1 MHz (medium wave, \lambda = 300 m) is a short monopole: half of a short dipole of length 2h, i.e. R_r = \frac{1}{2}\,20\pi^2\left(\frac{2h}{\lambda}\right)^2 = 40\pi^2\left(\frac{h}{\lambda}\right)^2 = 4.4 mΩ. With just 1 Ω of losses (loading coil, contacts, ground) the efficiency is 0.44 %. This is why medium-wave broadcast antennas are masts tens of metres tall, and why an AM radio in a car works only because the receiver has a lot of gain and external noise dominates anyway.

Efficiency of a short monopole with 1 ohm of losses
Efficiency grows with the square of the height relative to the wavelength as long as the radiation resistance stays small compared with the loss resistance: for the 1 metre whip at 1 MHz, red dot, it is 0.44 per cent.

The calculation in Python

import numpy as np
from scipy.integrate import quad

eta = 120 * np.pi                                    # impedance of free space (approximate) [ohm]

# cosine integral Cin(x) = integral from 0 to x of (1 - cos t)/t
Cin = lambda x: quad(lambda t: (1 - np.cos(t)) / t, 0, x)[0]
print(f"Cin(2 pi) = {Cin(2 * np.pi):.4f} -> R = 30 Cin(2 pi) = {30 * Cin(2 * np.pi):.2f} ohm")

def R_max(l_over_lambda):
    """Radiation resistance of a thin dipole of length l, referred to the current maximum."""
    kl2 = np.pi * l_over_lambda                      # k l / 2
    f = lambda th: (np.cos(kl2 * np.cos(th)) - np.cos(kl2)) ** 2 / np.sin(th)
    return eta / (2 * np.pi) * quad(f, 1e-9, np.pi - 1e-9, limit=200)[0]

for x in (0.01, 0.1, 0.25, 0.5, 1.0, 1.5):
    Rm = R_max(x)
    s2 = np.sin(np.pi * x) ** 2
    Rin = Rm / s2 if s2 > 1e-6 else float("inf")
    print(f"l = {x:4.2f} lambda: R (I max) = {Rm:8.3f} ohm, R input = {Rin:9.3f} ohm, short 20 pi^2 x^2 = {20 * np.pi**2 * x**2:.4f}")

# directivity of the half-wave dipole: D = 4 pi U_max / P = eta / (pi R)
R05 = R_max(0.5)
print(f"D (half-wave) = {eta / (np.pi * R05):.3f} = {10 * np.log10(eta / (np.pi * R05)):.2f} dBi; lambda/4 monopole: {R05 / 2:.2f} ohm")

# 1 m whip at 1 MHz (short monopole) and 1 ohm of loss resistance
h, lam = 1.0, 300.0
Rr = 40 * np.pi**2 * (h / lam) ** 2
print(f"1 m whip at 1 MHz: Rr = {Rr * 1e3:.2f} mohm, efficiency with 1 ohm of losses = {100 * Rr / (Rr + 1):.2f} %")

# ohmic losses of a copper half-wave dipole at 100 MHz, radius 1 mm (skin effect)
f0, a, sigma, mu0 = 100e6, 1e-3, 5.8e7, 4e-7 * np.pi
Rs = np.sqrt(np.pi * f0 * mu0 / sigma)               # surface resistance [ohm]
l = 3e8 / f0 / 2
R_unif = l * Rs / (2 * np.pi * a)                    # if the current were uniform
R_loss = R_unif / 2                                  # sinusoidal current: average of cos^2 = 1/2
print(f"Rs = {Rs * 1e3:.2f} mohm, loss R = {R_loss * 1e3:.1f} mohm, efficiency = {100 * R05 / (R05 + R_loss):.2f} %")

Running it prints \text{Cin}(2\pi) = 2.4377 and R = 73.13 Ω, the table of the previous section, D = 1.641 (2.15 dBi), 36.56 Ω for the monopole, 4.39 mΩ and 0.44 % for the whip, 311 mΩ and 99.58 % for the copper dipole. In MATLAB the same calculation is done with integral, and the Antenna Toolbox computes the complete impedance (including the reactance, not covered here) with the method of moments.

Subscribe
Notify of
guest
1 Comment
Oldest
Newest Most Voted
1
0
Would love your thoughts, please comment.x
()
x