Frequency Modulated (FM) radio has been used for years in high-fidelity music and speech broadcasting, offering excellent sound quality, signal robustness, and noise immunity. Recently, FM radio has witnessed an explosion of interest from the market for its applications in mobile and personal media players; however, the traditional FM design approach requires a long antenna, such as a wired headphone, which limits its usefulness for many users who do not carry the wired headset. Also, as wireless usage models continue to be a growing trend in portable devices, more customers can benefit from wire-free FM radio reception using other FM antennas while listening with either a wireless headset or a speaker output.
This article introduces an FM radio receiver solution that enables the antenna to be integrated or embedded inside the portable device enclosure, making the headphone cable optional. It starts with the goal of maximizing sensitivity; follows with methods for achieving the maximum sensitivity, including maximizing efficiency at the resonant frequency, maximizing antenna size, and maximizing efficiency across the FM band with a tunable matching network. Finally, this article describes an implementation of the tunable matching network.
Sensitivity can be defined as the weakest signal that an FM receiver system can receive while achieving a certain signal-to-noise ratio (SNR). It is an important parameter of FM receiving system performance and is related to both signal and noise. The received signal strength indicator (RSSI) indicates only the RF signal strength at a particular tuned frequency. It does not provide any information about noise or signal quality. The audio signal-to-noise ratio (SNR) is perhaps a better measure for comparing receiver performance with different antennas. Therefore, maximizing SNR is essential for listeners to experience good audio quality.
Antennas are the connection between the RF electrical circuits and electromagnetic waves. For FM reception, an antenna is a transducer that converts energy from electromagnetic waves to a voltage that can be used by an electrical circuit, such as a Low Noise Amplifier (LNA). The sensitivity of an FM receiving system is directly related to the electrical voltage received by the internal LNA. To maximize sensitivity, the electrical voltage must be maximized.
There is a variety of antennas, including headphone, stub, loop, and chip antennas, on the market, but all antennas can be analyzed using equivalent circuits. Figure 1 shows a generalized equivalent antenna circuit model:
In Figure 1, X can be either a capacitor or an inductor. The choice of X is determined by the antenna topology, where the value of the reactance (inductive or capacitive) is related to the antenna geometry. The loss resistance, Rloss, is related to the power dissipated in the antenna as thermal energy. The radiation resistance, Rrad is related to the voltage generated from the electromagnetic wave. For simplicity, we will analyze the loop antenna model in the remainder of this article. Similar calculations can be made for other antenna types, such as the short monopole and headphone antennas.
Maximizing Efficiency at the Resonant Frequency
In order to maximize energy from the antenna, a resonant network is used to cancel out the reactive impedance of the antenna, which would otherwise attenuate the amount of voltage the antenna transfers to the internal LNA. For an inductive loop antenna, a capacitor, Cres, is used to resonate the antenna at the desired frequency:
The resonant frequency is the frequency at which the antenna most efficiently converts an electromagnetic wave to voltage. The antenna efficiency is the ratio, of the power through Rrad to the total power collected by the antenna and can be written as Rrad / Zant, where Zant is the impedance of the antenna with the antenna resonance network. Zant is written as:
When the antenna is resonated, the efficiency, η, can be written as:
At other frequencies:
At frequencies other than the resonant frequency, ƒres, the antenna efficiency, η, is lower than the maximum efficiency, ηres, since the antenna input impedance, Zant, is either capacitive or inductive.
Maximizing Antenna Size
To recover a transmitted radio signal, the antenna must collect as much energy as possible from the electromagnetic wave and efficiently convert it into voltage through Rrad. The amount of energy collected is limited by the available space and size of antennas used in portable devices. For traditional headphone antennas, it can be as long as a quarter wavelength of the FM signal, which collects sufficient energy to convert to a voltage that can be used by the internal LNA. Consequently, it is less important to maximize the efficiency of the antenna.
Because portable devices are getting smaller and thinner, the space allowed for an embedded FM antenna is very limited. It is still important to maximize antenna size, but the energy collected by an embedded antenna is small. Therefore, to use smaller antennas without sacrificing performance, improving antenna efficiency, η, becomes very important.
Maximizing Efficiency across the FM Band with a Tunable Matching Network
In most countries, the FM broadcast band is in the frequency range of 87.5 to 108.0 MHz. In Japan, the FM broadcast band is 76 to 90 MHz, and, in some eastern European countries, the FM broadcast band is 65.8 to 74 MHz. To accommodate all FM bands worldwide, a 40 MHz bandwidth is required for an FM receive system. Traditional solutions usually tune the antenna at the center frequency in the FM band. However, as shown in the above equations, the efficiency of the antenna system is a function of frequency and reaches its maximum at the resonant frequency, dropping as the frequency is moved away from the resonant frequency. Again, since the worldwide FM band can be as wide as 40 MHz, antenna efficiency can decrease significantly at frequencies far from the resonant frequency.
For example, setting a fixed resonant frequency of 98 MHz gives good efficiency at this frequency point, but efficiency at other frequencies drops significantly, degrading FM performance the further one moves from the resonant frequency.
The graph below shows an efficiency plot for two antennas (a headphone antenna and a short antenna) with fixed resonance at the center of the band (98 MHz).
From the above graph, 98 MHz achieves the best efficiency, but the efficiency degrades closer to the band edges. This is not a significant issue for the headphone antenna since the antenna is large enough to collect sufficient electromagnetic energy to transfer a significant voltage to the RF receiver across the whole band; however, the short antenna is small and collects less energy compared to a longer headphone antenna, and the efficiency also rolls off faster as the frequency moves away from resonance. This can present a problem for reception at the band edges using fixed resonance. This is primarily due to the fact that a short antenna will likely have a higher "Q" than a headphone, resulting in the sharper drop at the band edges.
The quality factor, Q, is proportional to the energy stored in the antenna network to the energy lost or radiated, per unit time. For the above antenna equivalent circuit with an antenna resonated network, Q follows below:
A headphone antenna has inherently higher radiation resistance, Rrad, than a short antenna due to its larger geometry, resulting in a lower Q than the short antenna. The issue of efficiency roll-off is very pronounced with the short high-Q antennas required for embedded implementations.
The antenna's Q is also related to the bandwidth of the antenna. This relationship is given as:
where ƒc is the resonant frequency, and BW is the 3 dB bandwidth of the antenna. A short high-Q antenna has a smaller BW compared to a long headphone antenna and increases losses at the band edges.
To overcome the bandwidth limitations of a high-Q, fixed-resonance antenna, a self-tuning resonant circuit is used to change "fixed resonance" to "tuned resonance" so that the circuit is always at the resonant frequency for maximum sensitivity. A higher SNR is achieved with a self-tuning resonant antenna because the gain from the resonant antenna lowers the system noise figure of the receiver, and the inherent high Q of the embedded antenna helps filter interference that could mix with harmonics of the local oscillator.