In radio and telecommunications, a dipole antenna or doublet is one of the two simplest and most commonly used types of antenna. The other type is the monopole antenna. A dipole is part of a group of antennas that create a signal pattern similar to a basic electric dipole. This pattern is made possible by a structure that carries electrical current, with the current having only one point of minimum strength at each end. A dipole antenna usually has two equal parts made of conductive materials, such as metal wires or rods. The electrical current from a transmitter is sent between the two parts, or for receiving signals, the current is collected from the two parts. Each side of the cable connecting the transmitter or receiver is attached to one of the conductive parts. This is different from a monopole antenna, which has only one rod or conductor. One side of the cable connects to the monopole, and the other side connects to a ground. A common example of a dipole is the rabbit ears antenna used on television sets. All dipoles act like two monopoles placed together, with one connected to the other in a way that creates a balanced signal.
From a theoretical perspective, the dipole is the simplest type of antenna. It is often made of two equal-length conductors placed end-to-end, with the cable connected between them. Dipoles are often used as resonant antennas, meaning they work best at specific frequencies. If the connection point of the antenna is closed, it can vibrate at a certain frequency, like a guitar string that is plucked. Using the antenna near this frequency is helpful because it improves how well it sends or receives signals. The length of the antenna is chosen based on the wavelength (or frequency) it is designed to use. The most common type is the center-fed half-wave dipole, which is slightly shorter than half the length of the signal's wavelength. The signal from this dipole spreads strongest in directions perpendicular to the antenna and is weakest along its length. This makes it useful as an antenna that sends signals in all directions if placed vertically or as a slightly focused antenna if placed horizontally.
Dipoles can be used alone as simple, low-power antennas. They are also used as main parts in more complex designs, such as the Yagi antenna or driven arrays. Dipoles (or similar designs, like monopoles) are used to power more advanced directional antennas, such as horn antennas, parabolic reflectors, or corner reflectors. Engineers study vertical or monopole antennas by comparing them to dipole antennas, as monopoles are like half of a dipole.
History
In 1887, German physicist Heinrich Hertz showed that radio waves exist using a type of antenna now called a dipole antenna, which includes a feature known as capacitative end-loading. At the same time, Guglielmo Marconi discovered through experiments that he could connect the transmitter directly to the ground, eliminating the need for one half of the antenna. This design created a vertical or monopole antenna. For the low frequencies Marconi used to send signals over long distances, this type of antenna was more practical. However, as radio technology advanced to use higher frequencies, such as those used for FM radio and television (VHF), it became better to place the entire antenna on top of a tower, making dipole antennas or their variations more suitable.
During the early years of radio, the monopole antenna (also called the Marconi antenna) and the dipole antenna (also called the doublet) were considered separate inventions. Today, however, the monopole antenna is understood as a specific type of dipole antenna that includes an imaginary part underground.
Dipole variations
A short dipole is an antenna made of two conductors with a total length much shorter than half the wavelength (1/2 λ). These are used when a full half-wave dipole would be too large. Short dipoles can be studied using the Hertzian dipole, a theoretical model. Because they are shorter than a resonant antenna (half a wavelength long), their feedpoint impedance includes a large capacitive reactance. This means they need a loading coil or matching network to work well, especially as transmitting antennas.
To find the far-field electric and magnetic fields from a short dipole, we use results from the Hertzian dipole model. The radiator has a current of Iₕe^{jωt} over a short length ℓ. In electronics, j² = -1 replaces the usual symbol i for the square root of -1. ω is the angular frequency (ω = 2πf), and k is the wavenumber (k = 2π/λ). ζ₀ is the impedance of free space (ζ₀ ≈ 376.7 Ω), which is the ratio of the electric to magnetic field strength in free space.
The feedpoint is usually at the center of the dipole. The current along the dipole arms is proportional to sin(kz), where z is the distance to the nearest end. For a short dipole, the current drops linearly from I₀ at the feedpoint to zero at the ends. This is similar to a Hertzian dipole with an effective current Iₕ equal to the average current over the conductor, so Iₕ = (1/2)I₀. Using this, the equations for the fields generated by a short dipole fed by current I₀ are accurate.
The radiated flux (power per unit area) can be found using the Poynting vector, S = (1/2)E × H. Since E and H are perpendicular and in phase, the cross product simplifies to (1/2)EθHϕ. The phase factors cancel out, leaving the flux expressed in terms of the feedpoint current I₀ and the ratio of the dipole's length ℓ to the wavelength λ. The radiation pattern, given by sin²(θ), is similar to and slightly less directional than that of a half-wave dipole.
By integrating the radiation expression over all solid angles, we can find the total radiated power. From this, the radiation resistance (the resistive part of the feedpoint impedance) can be calculated, ignoring ohmic losses. Setting P_total equal to the power supplied at the feedpoint, (1/2)I₀²R_radiation, we find the radiation resistance.
These approximations are accurate when ℓ ≪ 1/2 λ. If ℓ = 1/2 λ (even though this is not strictly valid for such a large fraction of the wavelength), the formula predicts a radiation resistance of 49 Ω instead of the actual 73 Ω for a half-wave dipole, which uses more accurate quarter-wave sinusoidal currents.
The fundamental resonance of a thin linear conductor occurs when its free-space wavelength is twice its length (i.e., the conductor is 1/2 wavelength long). Dipole antennas are often used at this frequency and are called half-wave dipoles. This case is discussed in the next section.
Thin linear conductors are resonant at any integer multiple of a half-wavelength: ℓ = n(1/2)λ, where n is an integer, λ = c/f is the wavelength, and c is the reduced speed of radio waves in the conductor (c ≈ 97% of the speed of light). For a center-fed dipole, the driving point impedance differs significantly between odd and even multiples of half-wavelengths. Dipoles with odd multiples have low driving point impedance (purely resistive at resonance), while those with even multiples (integer wavelengths) have high driving point impedance (also purely resistive at resonance).
For example, a full-wave dipole can be made with two half-wavelength conductors end-to-end (ℓ ≈ λ). This provides about 2 dB more gain than a half-wave dipole. Full-wave dipoles are used in shortwave broadcasting with large effective diameters and high-impedance balanced lines. Cage dipoles are often used for this purpose.
A 5/4-wave dipole has a lower but not purely resistive feedpoint impedance, requiring a matching network. It has about 3 dB more gain than a half-wave dipole, the highest gain for any dipole of similar length.
Other dipole lengths do not offer advantages and are rarely used. However, half-wave dipoles can sometimes be used at odd multiples of their fundamental frequency. For example, amateur radio antennas designed for 7 MHz can also be used as 3/2-wave dipoles at 21 MHz. Similarly, VHF television antennas resonant at the low VHF band (65 MHz) are also resonant at the high VHF band (195 MHz).
A half-wave dipole consists of two quarter-wavelength conductors end-to-end (ℓ = 1/2 λ). The current distribution is a standing wave, approximately sinusoidal along the dipole, with nodes at the ends and an antinode (peak current) at the center (feedpoint). The current is described by I(z) = I₀ sin(kz), where k = 2π/λ and z ranges from -1/2 ℓ to +1/2 ℓ.
In the far field, the radiation pattern has an electric field given by Eθ = (I₀ ζ₀ k ℓ / 4πr) sin(θ) cos(π/2 cos θ). The directional factor cos[π/2 cos θ]/sin θ is nearly the same as sin θ for a short dipole, resulting in a similar radiation pattern.
A numerical integration of the radiated power (|Eθ|² / 2ζ₀) over all solid angles gives the total power P_total radiated by the dipole with peak current I₀. Dividing P_total by 4πr² provides the flux at a large distance, averaged over all directions. Dividing the peak flux (at θ = 0) by the average flux gives the directive gain of 1.64. This can also be calculated using the cosine integral. From this, the radiation resistance can be determined as done for the short dipole.
Dipole characteristics
The feedpoint impedance of a dipole antenna depends on its electrical length and where it is connected. Because of this, a dipole usually works best over a narrow range of frequencies. If the frequency moves outside this range, the impedance no longer matches well with the transmitter, receiver, or transmission line. The real (resistive) and imaginary (reactive) parts of the impedance, depending on the antenna's length, are shown in the graph. The calculations for these values are explained below. The reactance value depends strongly on the diameter of the conductors used; the graph shows results for conductors with a diameter of 0.001 wavelengths.
Dipoles much shorter than half the wavelength of the signal are called short dipoles. These have very low radiation resistance and high capacitive reactance, making them inefficient. More of the transmitter's energy is lost as heat due to the resistance of the conductors, which is greater than the radiation resistance. However, they can still be useful as receiving antennas for longer wavelengths.
Dipoles with lengths close to half the wavelength of the signal are called half-wave dipoles and are widely used. These have much higher radiation resistance, closer to the characteristic impedance of transmission lines, and their efficiency is nearly 100%. In general radio engineering, the term "dipole," if not specified, refers to a center-fed half-wave dipole.
A true half-wave dipole is one-half the wavelength (λ) in length, where λ = c/f in free space. This type of dipole has a feedpoint impedance of 73 Ω resistance and +43 Ω reactance, making it slightly inductive. To cancel this reactance and provide a pure resistance to the feedline, the dipole is shortened by a factor of k, resulting in a length of ℓ = kλ/2. The adjustment factor k depends on the conductor's diameter, as shown in the graph. For thin wires (diameter, 0.00001 wavelength), k is about 0.98, while for thick conductors (diameter, 0.008 wavelength), k is about 0.94. This happens because thinner conductors are more sensitive to changes in length, requiring smaller adjustments to cancel reactance. Thicker conductors have wider operating bandwidths because their reactance changes less with length.
For a typical k of 0.95, the corrected antenna length can be calculated using the formulas: length in meters = 143/f or length in feet = 468/f, where f is the frequency in megahertz.
Dipoles with lengths equal to any odd multiple of half the wavelength (1/2 λ, 3/2 λ, etc.) are also resonant, meaning they have little or no reactance. However, these are rarely used. A dipole with a length of 5/4 λ is more efficient in terms of power output and radiation direction. This antenna has a large negative reactance and requires an inductive matching network (like a tapped loading coil or antenna tuner) to function properly. It is desirable because it has the highest gain among dipoles that are not significantly longer.
A dipole is omnidirectional in the plane perpendicular to the wire, with radiation dropping to zero along the wire's axis. In a half-wave dipole, radiation is strongest perpendicular to the antenna and decreases as (sin θ)² to zero on the axis. Its 3D radiation pattern is roughly toroidal (doughnut-shaped), symmetric around the conductor. When mounted vertically, it radiates most strongly horizontally. When mounted horizontally, it radiates strongest at right angles to the conductor, with no radiation along the dipole itself.
Ignoring electrical losses, the antenna gain equals the directive gain. For a short dipole, this is 1.50 (1.76 dBi or -0.39 dBd). For a half-wave dipole, it increases to 1.64 (2.15 dBi or 0 dBd). For a 5/4-wave dipole, gain increases further to about 5.2 dBi, making this length desirable despite being off-resonance. Longer dipoles have multi-lobed radiation patterns with lower gain unless they are much longer. Other improvements, like adding a corner reflector or using dipole arrays, can increase directivity but are named separately.
Ideally, a half-wave dipole should be fed with a balanced transmission line matching its 65–70 Ω input impedance. Twin lead with similar impedance is available but rarely used because it does not match the balanced terminals of most receivers. Instead, 300 Ω twin lead is often used with a folded dipole. A folded dipole has a driving point impedance four times that of a simple dipole, closely matching 300 Ω. Older televisions and FM tuners often have 300 Ω balanced inputs. However, twin lead can be disturbed by nearby conductors, so care must be taken when using it for transmitting.
Many coaxial cables have a 75 Ω characteristic impedance, which would match a half-wave dipole. However, coax is a single-ended line, while a center-fed dipole requires a balanced line. Using coax creates an unbalanced line, causing unequal currents in the transmission line. This can make the transmission line act as an antenna itself, altering the radiation pattern and changing the impedance seen by the transmitter or receiver.
A balun is needed to connect coaxial cable to a dipole. A balun transfers power between the single-ended coax and the balanced dipole, sometimes changing the impedance. It can be made with a transformer on a ferrite toroidal core. The core material must be suitable for the frequency and large enough to avoid saturation in transmitting antennas. Other balun designs include current baluns, which use a transformer on a magnetic core to block unwanted currents, and designs with two transformers that also provide impedance changes.
Common applications
Dipole antennas are widely used in many applications. One common example is the "rabbit ears" or "bunny ears" television antenna, which sits on top of broadcast TV receivers. This antenna receives VHF terrestrial TV signals in the United States, which include two frequency ranges: 54–88 MHz (Band I) and 174–216 MHz (Band III). These frequencies correspond to wavelengths of 5.5–1.4 meters (18 feet to 4 feet 8 inches). Because this wide range of frequencies cannot be covered by a single fixed dipole antenna, the rabbit ears design includes adjustable parts. It uses two telescoping rods that can extend to about 1 meter (3 feet) in length, which is one-quarter the wavelength at 75 MHz. By adjusting the length, angle, and direction of the rods, users can improve signal reception more effectively than with a rooftop antenna, even one with a rotating mechanism.
In contrast, the FM radio broadcast band (88–108 MHz) is narrow enough for a single dipole antenna to cover. For home use, hi-fi tuners often come with simple folded dipole antennas that are tuned to the center of this band. A folded dipole has four times the impedance of a simple dipole, which matches well with 300 Ω twin lead wiring. This type of antenna is often made using twin lead wire itself, with the ends connected together. It can be easily attached to walls or moldings, making it flexible for installation.
Horizontal wire dipole antennas are popular for use on HF (shortwave) bands, both for transmitting and receiving signals. These antennas are usually made of two wires joined at the center by a strain insulator, which serves as the feedpoint. The ends of the wires can be attached to buildings, trees, or other structures to take advantage of their height. When used for transmitting, it is important to connect the ends to supports using strain insulators with high flashover voltage to prevent electrical sparks, as high-voltage points occur at the ends. Since these are balanced antennas, they are best connected to a balun, which links the coaxial cable to the feedpoint.
These antennas are simple to set up for temporary or field use. They are also widely used by radio amateurs and shortwave listeners in fixed locations because of their low cost and ease of construction. They work well at frequencies where resonant antenna elements need to be large, but they are still practical for use when directionality is not required. Building multiple antennas for different frequency bands can be less expensive than buying a single commercial antenna.
Antennas for MF and LF radio stations are often built as mast radiators, where the vertical mast itself acts as the antenna. While most mast radiators are monopoles, some are designed as dipoles. In these cases, the metal mast is split at its midpoint into two insulated sections, creating a vertical dipole that is connected at the center.
Many array antennas are made using multiple dipoles, typically half-wave dipoles. The purpose of using multiple dipoles is to increase the antenna's directional gain compared to a single dipole. The signals from the dipoles interfere constructively in desired directions, boosting power output. In arrays with multiple driven elements, the feedline is split using an electrical network to provide power to the elements, with careful attention to the phase delays between the common point and each element.
To increase antenna gain in horizontal directions (while reducing radiation toward the sky or ground), antennas can be stacked vertically in a broadside array, where they are fed in phase. This setup maintains the dipoles' directionality and nulls in the direction of their elements. However, if the dipoles are vertically oriented in a collinear array, the null direction becomes vertical, and the array produces an omnidirectional pattern in the horizontal plane. These arrays are used in VHF and UHF bands, where the small size of the elements allows multiple antennas to be stacked on a mast. They are a higher-gain alternative to quarter-wave ground plane antennas used in fixed base stations for mobile two-way radios, such as police or taxi dispatchers.
For a rotating antenna or one used in a specific direction, increased gain and directivity in a single horizontal direction may be desired. If the broadside array is turned horizontal, the gain increases in the direction perpendicular to the antennas, but this also creates high gain in the opposite direction. To reduce this, a large planar reflector can be used, as in a reflective array antenna, to redirect energy and increase gain in the desired direction by an additional 3 dB.
Another type of directional antenna is the end-fire array. In this design, dipoles are placed side by side but not collinear, and they are fed with progressive phase shifts. This causes the signals to add coherently in one direction while canceling in the opposite direction. Unlike a broadside array, the directivity is along the line connecting the feedpoints, with the opposite direction suppressed.
Antennas with multiple driven elements require complex systems for splitting signals, managing phase delays, distributing power, and matching impedance. A simpler alternative is the Yagi antenna, which uses parasitic elements. In a Yagi antenna, only one dipole is connected to the feedline, while the others act as passive elements that reradiate energy. The lengths and positions of these parasitic elements are carefully chosen to focus gain in one direction and cancel radiation in the opposite direction. While the gain is slightly less than a driven array with the same number of elements, the Yagi antenna is more practical for consumer use due to its simpler electrical connections.
Antenna gain is often measured in decibels relative to a half-wave dipole. This is because practical measurements require a reference point, and the half-wave dipole is well understood and can be made nearly 100% efficient. The gain of a dipole is considered "free" since it is inherently directional. When gain is measured relative to a dipole, it is expressed as "x dBd." However, gains are more commonly expressed relative to an isotropic radiator, which appears to have no directionality. Since a half-wave dipole has a gain of 2.15 dBi, any gain in "dBi" is 2.15 dB higher than the same gain in "dBd."
Hertzian dipole
The Hertzian dipole is a theoretical model used in physics to study how antennas work. It represents a very small piece of conductor that carries a radio frequency (RF) current. This current has the same strength and direction along the entire short length of the conductor. Real antennas can be understood by combining many of these tiny Hertzian dipoles placed together end-to-end.
A Hertzian dipole can be described as a small segment of conductor with a current that changes over time in a specific pattern. This current flows in one direction and has a length so small it is almost invisible. By calculating the electromagnetic fields produced by a Hertzian dipole, scientists can predict how more complex antennas, like practical dipoles, radiate energy. This is done by combining the effects of many Hertzian dipoles that make up the current pattern of the real antenna. The total field is calculated as an integral over the path of the antenna conductor, which is modeled as a thin wire.
To simplify the calculations, the current is assumed to flow in the z-direction, centered at the origin (where x, y, and z are all zero). The time dependence of the current is described by a sine wave pattern, represented as e^{iωt}. Using the formula for the vector potential A(r), which describes the magnetic field, scientists calculate the fields produced by the dipole. The Lorenz gauge is used to simplify the equations, and the wave number k is defined as k = ω/c, where ω is the radian frequency and c is the speed of light in free space. The distance from the origin to the point being studied is denoted as r, which is the magnitude of the position vector r.
The vector potential A(r) is found to be purely in the z-direction, matching the direction of the current. Using this, the magnetic field H and the electric field E are calculated. In spherical coordinates, the magnetic field H has only a component in the ϕ direction, while the electric field E has components in both the θ and r directions. The impedance of free space, ζ₀, is defined as the square root of the ratio of the permeability of free space (μ₀) to the permittivity of free space (ε₀).
The solution includes near-field terms, which are strong close to the source but do not contribute to radiation. Near the source, the electric and magnetic fields are nearly 90° out of phase, reducing their contribution to the Poynting vector, which measures radiated power. The near-field solution is used to calculate mutual impedance between nearby antenna elements.
For far-field radiation, only the terms that decrease as 1/r remain significant. In the far field, the electromagnetic wave is transverse, meaning the electric and magnetic fields are perpendicular to each other and to the direction of propagation (the r direction). The electric field is polarized in the θ direction, which is coplanar with the current in the z direction, while the magnetic field is in the ϕ direction. At these distances, the fields are in phase, and both decrease as 1/r. The power radiated decreases as 1/r², following the inverse square law.
If the far-field radiation pattern of an antenna is known, the radiation resistance can be calculated directly. For the Hertzian dipole, the average power radiated is found using the Poynting vector. As the distance r increases, the radial component of the Poynting vector becomes dominant. The power per unit area crossing a large sphere around the source is calculated, and integrating this over the sphere gives the total radiated power. The radiation resistance R_rad is determined by equating the total power to (1/2)|I|²R_rad, where |I| is the current amplitude.
This method can be used to compute the radiation resistance for any antenna with a known far-field pattern. If ohmic losses are ignored, the radiation resistance is equal to the resistive part of the feedpoint impedance. However, this does not provide information about the reactive (imaginary) component of the impedance, which requires further analysis.
The directive gain of the Hertzian dipole can also be calculated using the Poynting vector. By dividing the total radiated power by the surface area of a sphere (4πr²), the average flux P_avg is found. Dividing the flux in a specific direction by P_avg gives the directive gain G(θ). The peak gain of the Hertzian dipole is between 1.5 and 1.76 dBi, which is lower than most other antenna configurations.
The Hertzian dipole is similar to the short dipole, but they differ in their current distribution and standing wave patterns. Both have conductors much shorter than a wavelength, but their behavior and calculations vary slightly.
Detailed calculation of dipole feedpoint impedance
The impedance at the feedpoint of a dipole antenna of different lengths is shown in a graph, using two parts: the resistive part (R dipole) and the reactive part (j X dipole). When an antenna has perfect conductors (no energy loss as heat), the resistive part (R dipole) is the same as the radiation resistance. This resistance can be calculated using the total power radiated in the far field, as explained for a short dipole. Calculating the reactive part (X dipole) is more complex.
Using the induced EMF method, direct mathematical formulas are found for both parts of the feedpoint impedance. These formulas depend on assuming a specific pattern for the current along the antenna. For antennas where the ratio of wavelength to element diameter is greater than about 60, the current along each half of the dipole (length 1/2 L) closely follows a sine wave pattern, with the current reaching zero at the ends (z = ±1/2 L).
In this case, the wavenumber (k) is calculated as k = 2π/λ = 2πf/c, where λ is the wavelength, f is the frequency, and c is the speed of light. The amplitude (A) is adjusted to match the current at the center of the antenna (z = 0).
When the current distribution is approximately sinusoidal, the driving point impedance can be calculated using cosine and sine integral functions (Si(x) and Ci(x)). For a dipole of total length L, the resistive and reactive parts of the impedance depend on the conductor radius (a), the wavenumber (k), the impedance of empty space (ζ₀ ≈ 377 Ω), and Euler’s constant (γₑ ≈ 0.5772). Some authors use an alternate form involving a different function called Cin.
The induced EMF method assumes a sinusoidal current pattern and is accurate to within about 10% when the wavelength-to-element diameter ratio is greater than 60. For larger conductors, numerical solutions are needed to calculate the current distribution without assuming a sine wave. These solutions use equations like Pocklington’s integro-differential equation or Hallén’s integral equation, which apply to a wider range of conductor shapes.
Numerical solutions use the moment method, which requires breaking the conductor into segments and approximating the current with simple functions. A matrix (N × N) is created and inverted to find the solution. Each matrix element involves complex calculations, but using delta functions as weighting functions simplifies the process by focusing on discrete points along the conductor. As the number of segments (N) increases, the matrix inversion becomes more computationally demanding. For example, Balanis (2011) used Pocklington’s method and found that solutions stabilize to within a few percent when N exceeds 60.