In the simplest cell phones, the ringtone is a buzz or a repeated ring resembling that of an old landline phone. In an entertainment center type, the ringtone is often an excerpt from a musical composition. Also, continuous music may be played through it. As shown, the handsfree loudspeaker opening is on the side of the cell phone so that the phone can be laid on a flat surface and thus is “handsfree.” If the surface is large enough, the energy output of the loudspeaker will be double that compared with the handheld position because the opening is radiating into half space. Like any small closed-box loudspeaker the trend in cell phones is toward the “acoustic suspension” concept, i.e., the stiffness of the suspension for the diaphragm is a small fraction of that of the airspace behind. The volume of the backspace is about 1   cm³. This design is only practical if the diaphragm is fairly stable and the resonant frequencies are constant. Surprisingly, in the cell phone of Fig. 8.1, the user can carry on a telephone conversation with it on a surface, despite the proximity of the microphone to the handsfree opening. This is possible because the output of the microphone has an “echo canceling” feature in the digital circuitry and its output is also attenuated while the loudspeaker is operating. Nowadays handsfree loudspeakers are very loud, often producing as much as 105   phons at a distance of 10   cm.

The receiver is similar to the handsfree loudspeaker. In Fig. 8.1, the volume of the space behind its diaphragm is larger and it contains the electronics and the keyboard. There is greater chance of leakage from this space and, if it should occur, the low-frequency output will be diminished. The opening is always held close to the ear, which mitigates the effect of leakage from the rear space.

The microphone is an electret or microelectromechanical system (MEMS) type, more of which will be discussed later. It is only actuated when voice transmission is required.

In cell phones, for both the handsfree loudspeaker and the receiver, the compliance of the front cavity C _AF and the mass of the sound opening M _AH form a Helmholtz resonator. This is a second-order low-pass filter. The angular resonance frequency is

All of the openings, for the loudspeakers and microphone, must be covered by a dust screen, which not only protects against the ingress of magnetic dust but also helps damp out the various Helmholtz resonances. Data on dust screens are given in Section 4.4.

High strength neodymium magnets have generally improved the performance of miniature loudspeakers; however, their proximity to sound holes means that magnetic dust can be sucked in and clog the coil gap. Magnetic fields can also affect magnetic strips on credit cards.

The location of a loudspeaker opening is very important because, if it is tightly held against the ear, damage to hearing may result. Loudspeaker openings are often located on the sides of a phone so that they cannot be sealed against the ear. Another frequent location is on a rear surface that is curved, which also prevents closure of the opening when laid on a table.

It is vital that all seals be secure. If the back enclosure leaks, the stiff cushion of the air that limits the excursion of the diaphragm no longer exists and physical damage may result. At low frequencies, the generated sound pressure is limited by diaphragm excursion and at high frequencies by power dissipation. The input voltage to a voice coil is often limited according to a digital signal processor model [2] of the loudspeaker parameters and how they change according to conditions such as temperature. In addition, audio peak clipping, called dynamic range compression, is used extensively to make the sound louder.

At low particle velocities, the flow of air in a tube is linear. In other words, the velocity increases with radial distance from the wall as discussed in Chapter 4. As the velocity is increased beyond a certain point, the flow becomes turbulent with the formation of chaotic eddies and vortices. The Reynolds number is given by

where

The location of the microphone on the handset is critical with respect to wind noise. The worst location is on a large flat area such as a front cover because there is greater boundary layer turbulence than on a small area. The bottom edge is better and the best position is away from the center of the bottom edge. The direction of the wind also makes a significant difference. When the microphone is on the surface facing the wind, there is generally less noise than when it is facing away because turbulence builds up as air flows past an obstacle. Trailing edge vortex shedding also contributes to wind noise, but to a lesser degree than boundary layer turbulence. In general, all surfaces should be as smooth as possible with no sharp edges.

Handling noise is generated by friction when a handset is held. Smooth surfaces produce less noise than rough ones. The noise is amplified by structural vibrations within housing as well as normal modes within the internal space, although the latter will be reduced by the presence of electronic components. Therefore, the housing should be either rigid or well damped, or both. Adding ribs to the inner surfaces of the housing increases its rigidity. These precautions will also help to reduce unwanted resonance peaks and dips in the handsfree loudspeaker and receiver frequency responses.

Until the electret condenser microphone (ECM) was invented by Sessler and West [5–8], all condenser microphones required a polarizing voltage supply. Such condenser microphones are treated in Section 5.5. The name electret literally means “electrostatic magnet.” It is a dielectric in which a permanent charge is embedded. In Sessler and West's invention, as shown in Fig. 8.5(a), the dielectric was a metalized Teflon foil which formed the diaphragm and the electrode was a stationary plate located behind it. In recent years, the charge storage capacity has been increased by using porous membranes [9]. Metalized electrets have a dipole charge because a charge of opposite polarity to that contained by the membrane is induced in the metallic coating. Although this gives a more stable charge, it means that, for a given membrane charge, the resulting field decays with increasing electrode separation, whereas in an externally polarized condenser microphone, the field remains constant. However, this is not a major issue in a microphone where the electrode spacing is usually very small. The capacitance of the electret membrane is given by

where ε ₀ is the permittivity of air, ε _r is the relative permittivity of the dielectric, S is the area of the membrane, and h is the thickness of the electret membrane. This voltage then also appears across the air gap when the microphone terminals are at the same potential. The capacitance of the air gap is given by

where d is the width of the gap. Then the total capacitance across the input terminals is

so that when d   =   ∞, there is zero charge on the electrode (σ ₀   =   0) and the charge on the coating is equal and opposite to the electret charge so that σ ₁   =   σ _m . In general, the total induced charge on the electrode and coating is equal and opposite to the electret charge. When d   =   0, the induced charge is shared equally so that σ ₀   =   σ ₁   =   σ _m /2. Although the charge may be distributed throughout the electret, we may model it as a concentrated layer somewhere near the middle, depending on how symmetrical the charge distribution is. Inevitably some charge is lost both near the outer surface and near the coating, where there will be some recombination of positive and negative charge. Hence we may treat the concentrated charge layer and coating as electrodes of a capacitor across which there is a polarizing voltage E, where the dielectric thickness is h/2, so that

At the time of writing, most low-cost cell phones use analogue electret microphones, whereas midmarket models use both analogue microelectromechanical system microphones and digital electret microphones, where the latter are analogue electret microphones with onboard analogue-to-digital converters. This has the advantage of immunity to electrical noise pickup by the connections between the microphone and baseband chip. Many top-end models now use digital MEMS microphones because these have better signal-to-noise ratios (typically 69   dB). Also, MEMS microphones are smaller and can be surface mounted, which is an advantage in products that may have stereo microphones for recording as well as multiple microphones for noise cancellation. The sensitivity of a MEMS microphone is typically −38 to −42   dBV/Pa, and the acoustic performance is very stable.

Because there are many measurements in type approval testing, we shall only discuss the basic acoustic ones.

Here we shall confine ourselves to broadband telephony (100   Hz–8   kHz) as opposed to narrowband (100   Hz–4   kHz). To measure the send frequency response, the sound source is calibrated by means of a microphone holder attached to the head and torso simulator (see Fig. 8.4). Then the calibration microphone is attached to the microphone holder with the diaphragm at the mouth reference point (MRP), which is 25   mm in front of the lip plane. The sound source is equalized to give a constant sound pressure of typically −4.7   dB   Pa at all test frequencies. The frequency response is measured by connecting an analyzer to the point of interconnection (POI). In a landline phone, this is an analogue connection to the public switched telephone network, but in a cell phone it is usually the digital interface point (formerly digital air interface). The measured frequency response is plotted between upper and lower limits known as a mask. Thus, to meet the requirements of type approval, the plot must fall within this mask. A typical frequency response is shown in Fig. 8.11.

The loudness rating LR is calculated from the measured response over a set of frequencies which are spaced 1/3 octave apart according to the following formula:

where

When we talk, we normally hear our own voices via the external acoustic path between our mouths and ears. It is unnatural to have one ear obscured by a cell phone, so the acoustic path is replaced by an electrical one between the microphone and receiver. This is known as sidetone. There are two kinds, one being a deliberate path within the handset and the other the result of echo from an imperfect analogue interface, or hybrid, to a two-wire line within the network. The latter is minimized through use of a digital echo canceler because delayed sidetone is extremely irritating to the caller. Other sidetone paths may be carried by the mechanical structure of the phone or the space within it.

where D is a parameter of the phone which is independent of the network.