Any conductive body functions intentionally or otherwise as an antenna in the presence of an oscillating electromagnetic field. This radiation excites free electrons in the conductor, and if a load is connected, a small but measurable current will flow in that load. By the same token, if oscillating voltage is applied to a conductor, it will emit electromagnetic radiation into the surrounding space. Proper orientation and impedance matching greatly enhance this effect. Sizing with respect to the wavelength is critical.

Whether implemented for transmission or reception, an antenna’s electrical qualities are the same. This is a consequence of the reciprocity theorem, and it is applicable to parameters such as radiation pattern, impedance, gain, bandwidth, polarization and resonant frequency. The design of an antenna determines its directional qualities, radiation pattern and gain. An ideal spherical transmitting antenna would radiate electromagnetic radiation equally in all directions, although this is never realized because the transmission line has to be connected somewhere. Moreover, the ideal transmitting antenna would be wasteful and undesired.

Transmitting and receiving antennas can be made slightly or highly directional by means of their orientation. For example, they should not direct transmitted power into the ground nor upward unless intended to communicate with aircraft or satellite, or as radar. Moreover, transmitting antennas generally direct broadcasts toward more populous areas and not over the ocean.

The radiation pattern results in antenna gain for the area targeted. This gain has nothing to do with the gain of an amplifier, where amplification provides the gain. The gain of an antenna is a consequence of directionality. Power that is not transmitted in an undesired direction goes into increased power directed elsewhere. Point-to-point, as opposed to broadcast transmission, is highly directional and has high gain when both antennas are situated at the foci of parabolic dishes, as in repeater stations. Satellite dish antennas, used in areas where cable is not available, must be precisely aimed at satellites that are in geostationary orbit. Precise aiming is more critical for Internet than TV communication.

Gain also depends upon the antenna’s efficiency and the precision with which it is manufactured and installed with respect to the carrier frequency. Frequency (f) and wavelength (λ) are in a reciprocal relationship:

f = 1/λ and λ = 1/f.

Conventional antenna design is largely based on the wavelength or range of wavelengths of the carrier signal. (Wavelength refers to the linear distance through space occupied by one complete cycle, both negative and positive portions).

The monopole antenna, for example an automotive whip antenna, is horizontally omnidirectional. It is connected to a single transmission line conductor. The return current path is ground, either the earth or a large conductive body such as an aircraft that comprises a virtual ground.

Other than the monopole design, just about all antennas are dipole in one form or another. The half-wave dipole consists of two quarter-wavelength metal conductors placed end-to-end so they lie along the same axis. The half-wave dipole is a resonant antenna that depends upon the principle that when electrons flowing through a conductor encounter an impedance mismatch, some or all of the electrons at the joint will be reflected back toward the source. So when the oscillating current gets to the free ends of the two dipole elements, it is reflected back, and if the impedance is then a match, it moves through the source.

If the dipole antenna is exactly a half-wavelength long, each half element being a quarter-wavelength long, the antenna resonates with the electromagnetic energy as transmitted. That is because at one end the electromagnetic energy and the tip of the antenna are simultaneously at maximum positive value while at the other end they are at maximum negative value. Thus a standing wave is created along the dipole antenna. This phenomenon also depends upon the dipole being exactly aligned with the carrier wave direction of propagation.

These types of harmonically operated antennas can also be constructed as one-third the length of the wave at the design frequency. Other odd multiples are viable because they are odd harmonics. (Even harmonics cancel with the fundamental.)
The design and installation of the transmission line, from antenna to transmitter or receiver, is critical. The important parameter is characteristic impedance, which must match the input and output impedances to avoid reflections. Also, the terminations must be designed, installed and maintained to avoid any discontinuities.

A balun can be placed in the circuit for impedance matching or to interface balanced and unbalanced circuits. A balun is a small transformer or choke used to connect source-to-line or line-to-load. Baluns are made of diverse materials, with and without ferrite cores and internal connections between windings. They are configured in different ways so as to serve various purposes. For such a simple little device, the balun is a complex and engrossing subject.

An antenna can be made to operate at a frequency that is lower than its actual resonant frequency. A monopole or dipole antenna can be manufactured shorter than a quarter or half-wavelength or at odd multiples. Such an antenna is known as an electrically short antenna. Its impedance is then highly influenced by capacitive reactance that is effectively in series with the antenna. A loading coil, which contributes series inductance, can be inserted into the circuit and it will have the effect of canceling the capacitive reactance so the remaining impedance is resistive only and the resonant frequency is lowered as if a longer antenna had been used.

In a receiving antenna, when the source is weak due to distance or transmitter qualities, the first thought might be to make the antenna elements larger. The problem here is that we are moving farther from a resonant condition. One widely-used remedy is to build an antenna array, consisting of elements that are not electrically connected to the main electrode or to the transmission line.

To understand how an array works, consider one antenna that acquires a modulated standing wave. The electrical energy in the conductive antenna is oscillating and in turn, is emitted as an oscillating signal. So the receiving antenna also functions as a transmitting antenna. (Any conductive body does this.) Additional nearby elements that are shaped, sized, located and oriented properly will intercept and re-radiate the carrier wave received by the main antenna, reinforcing the standing wave. Multiple passive elements can be deployed in this fashion, strengthening the signal at the receiver.

The Yagi-Uda array once was a common design back when most households got TV signals over the air. On a tall vertical pole, antenna elements were oriented broadside to the signal for maximum reception. One element was active, electrically connected to the transmission line. The other elements were passive. The elements in front of the active element were called directors and those behind were reflectors. This terminology explains how they worked.

Adding more elements made the Yagi-Uda array more directive and boosted its gain. It had a higher Q factor and thus lower bandwidth. Consequently, it was more frequency sensitive and performance would decline in heavy winds or if aging roof framing moved a slight amount.

The log periodic antenna has a similar appearance, but all elements are active, connected to a transmission line with segments sized to specific lengths to produce an in-phase signal.

Antennas designed for cellular or Industrial-Scientific-Medical (ISM) band applications below 900 MHz are physically small. So the size of the printed-circuit-board (PCB) to which they attach can drastically affect their average gain and efficiency. A complicating factor is that flex antennas and their feed lines often employ radiation from the feedline as part of the antenna’s total radiation pattern. So the physical position of the feed line can affect the performance and antenna patterns produced by a flex antenna on a particular PCB design.

Transmitters and receivers operating near ISM band frequencies often employ chip antennas. Chip antennas employ a dielectric or cavity resonator. The cavity between the conductors is comprised of a material having a high permittivity constant, often a ceramic. Use of a high ϵ_r bulk material lets dielectric resonator antennas be significantly smaller than the wavelength of the frequency of interest in free space (at 300 MHz, λ = 100 cm; at 5 GHz, λ = 6 cm).

Chip antennas generally have radiation patterns resembling that of classical dipole antennas. But chip antenna suppliers typically test antennas on an optimized demo board. Actual applications typically include clutter from surrounding structures such as batteries, cables, and displays. These obstacles can degrade the performance of a PCB-mounted antenna in ways that won’t show up when the antenna is tested alone.

Products containing transmitters or receivers must undergo FCC/EN radio compliance testing. Compliance testing focuses on the interference that the radio design may generate. It involves checking for spurious emissions and examining band edges and modulation parameters in detail. A point to note is that such testing does not attempt to quantify the antenna performance. For example, an FCC compliance test won’t reveal whether an antenna is inefficient or exhibits unintentional directivity.

The most common antenna measurements generally include the radiation pattern, antenna gain and efficiency, antenna impedance or VSWR, bandwidth, and polarization. Antenna performance is often assumed to be good when an antenna shows a low VSWR (minimum VSWR = 1) or minimal return loss when tested with a vector network analyzer. But antennas with return-loss below 10 dB can also suffer from poor radiation efficiency (less than 10%). Trial-and-error impedance matching with inductive/capacitive impedance-matching components may help this situation.

In contrast, antenna circuits with VSWRs above 10:1 or more can experience a large amplification of current in impedance-matching inductors and a voltage boost in impedance-matching capacitors. These conditions multiply the inherent losses of these components. The high losses masquerade as radiation resistance, though the antenna may not be practically matchable. Thus the matching-network components may seem to yield a “good match” by absorbing RF energy, not by transferring it to a radiating antenna.

To detect such difficulties, engineers might employ a “hand test” by waving their hand over a board and its antenna and noting whether this action affects return loss measurements. If the VSWR doesn’t change, then the design is acting as a dummy load rather than as an antenna. In addition, a compact antenna that exhibits unexpectedly wide return-loss bandwidth probably does so as a result of high internal losses and not the radiation resistance.

An additional point to note is that any structures within a quarter wavelength of the antenna must be treated as being part of the antenna. They will absorb RF energy and lower the antenna’s efficiency while also reflecting energy and causing changes to the antenna radiation pattern.

The best way to check for such difficulties is on an antenna range as maintained by test labs. Most such tests take place in an anechoic chamber. Usual test parameters include gain; radiation efficiency; VSWR (return loss); radiation patterns in boresight, polar, or spherical formats (1D, 2D, or 3D formats); left-hand-circular-polarization (LHCP) gain; right-hand circular polarization (RHCP) gain; and axial ratio.