1
General Review

Speech and music is noise with meaning. The recording and reproduction of sound is imperfect, and the imperfections in these processes reduce meaning and add noise. The art of the loudspeaker designer is the employment of science to help increase meaning for reproduced sound. An understanding and familiarity with music and live sound is fundamental for a reasoned application of acoustical engineering to the still-imperfect process of loudspeaker design. Great advances have been made with materials technology, refined acoustical modeling, electrical theory and design software, all these helping to manipulate, control and balance all the factors involved in the engineering and construction, but there remains a substantial human component, namely the subjective judgment of reproduced sound quality, which is a skill the loudspeaker designer must acquire.

Some design engineers may view speakers as acoustical machinery, and thus design and build them as such. Conversely, loudspeakers must be considered as imperfect reproducers of speech and music. If you know and love music, then the machinery design part of loudspeakers should be guided by a continual assessment of the quality of the musical experiences that they create. Here science must serve art.

Reference 'monitor' speakers may be designed which are highly competent mechanically and acoustically, are usefully informative, but which may not fully inspire the listener with the quality of their musical performances. Highly qualified and trained engineers may well be responsible for numerous notionally accurate loudspeakers while some engineers are not particularly musical. Often they imagine that application of pure science, 'done by the book', will be sufficient to complete the work.

1.1 Early Loudspeakers

It is some 90 years since the ubiquitous moving-coil loudspeaker was first developed as we know it, the mass-controlled, paper cone direct radiator: an electrodynamic transducer which converts electrical current into sound pressure at a useful loudness. In contrast to most other sound transducers it possesses an intrinsically uniform frequency response. It is clearly highly reliable in use, and comes with the proven potential for economic manufacture.

Before this development there were numerous 'earphone' type transducers, moving iron and the like of various kinds for music reproduction, some 'amplified' by improved coupling to the air with various early horn configurations. Certainly, somewhat earlier, primitive forms of moving coil and diaphragm sound signal reproducers had been made. Back in 1874, a U.S. patent for Siemens by Ernst Werner, was one of these, though at that time no electrical audio signals were available to drive it and so it was never heard to reproduce sound, instead emitting pulsed signal noises. And certainly, Peter L Jensen working with Pridham, (Figure 1.1) had developed a significantly powerful horn loaded loudspeaker by 1914, where the transducer employed a 75?mm diameter diaphragm of nickel silver alloy and employed electromagnetic field excitation acting on the moving coil. It was used successfully for large-scale public addresses for many years.

Figure 1.1 A horn driver from the Pridham Patent, issued 1923, showing a moving coil and an edge clamped planar diaphragm.

However, the familiar mass-controlled direct radiator moving-coil cone loudspeaker, whose principle is so effective that its key elements have remained essentially unchanged to this day, came with 'the New Hornless Loudspeaker' of 1925 by Rice and Kellogg of GE (USA). This set the stage for the, low-resonant frequency, direct radiating type of drive unit we know so well, a driver where a good part of the primary frequency response is intrinsically uniform with frequency, and which also may be acoustically loaded at lower frequencies to usefully extend the working range, by controlling the potential for front (positive) to back (negative) radiation cancellation from this intrinsically dipolar transducer.

1.1.1 The Elements of the Ubiquitous Cone Loudspeaker

To build such a transducer, take an affordable magnet and incorporate a simple arrangement of magnetically permeable 'soft' iron to help concentrate much of the available magnetic flux into a narrow radial gap using a cylindrical, central magnetic pole. A small light coil, a 'solenoid', is wound on a low mass former. This can be a tube of thin card. The assembly is suspended freely in the magnetic gap using a radially corrugated flexible disc or similar, allowing axial motion of a quarter centimeter or so.

In accordance with Maxwell's electromagnetic equations, an axial force is generated on the coil when current flows through it. This force is the product of B, the magnetic field strength, l the length of the wire immersed in that flux field and I, the current flowing through the coil. This force coupled relationship is fundamentally linear and consequently there is very little inherent distortion.

It is intrinsic for a moving-coil motor that there is effectively no lower resolution limit for small signals. An infinitely small electrical input will produce an equivalent and essentially infinitely small sound output. Another excellent feature of the moving-coil transducer, generally taken for granted, is that despite its nature as a moving mechanical device, it is essentially noiseless. It does not grate, scrape or whirr. Apply a sub-audible 5?Hz sine-wave current and you can see the coil move, but silently. The moving coil used on its own generates almost zero sound output. Radiated sound level is proportional to the area of air load driven or coupled by the transducer element, and for the coil alone, it comprises a thin ring of negligible radiating area.

It is essential to couple this moving element to a larger air load and thus a rigid, light diaphragm, generally of much larger area, is securely bonded to the coil former. Typically, such larger diaphragms have their own flexible outer surround, constituting a second suspension, fixed to an, skeletal, non-reflective support frame or chassis, the assembly providing the vitally important axial centering of the moving system, which can now be positioned in a close tolerance magnet gap. Such fine tolerance helps increase the magnetic flux density in the gap so maximizing efficiency and thus loudness.

We know that flat paper sheet is desirably lightweight, but it is very weak in bending. However, paper is remarkably stiff in tension. To make use of the latter property simply cut out some suitable paper in a useful shape, curl it up into a cone and glue the remaining seam. This simple conical structure exhibits an extraordinary axial stiffness for its mass, a marvellous means of coupling a much larger area of air load to that otherwise near silent moving-coil motor, so aiding conversion efficiency of force into sound. The coil former is firmly glued to the cone apex.

Here acting as an acoustical transformer, the cone or diaphragm matches the much lower acoustical impedance of the air load to the higher driving force impedance of the coil assembly, thus greatly improving the efficiency of energy transfer from electrical power, now providing readily audible sound pressure.

This signal path includes the translation of electric audio currents to mechanical forces; these are then coupled to a cone to usefully radiate sound pressure to be heard at a distance. The result is the familiar loudspeaker. This elegant principle has proved highly useful and effective for almost a century, even if in absolute terms the conversion efficiency from input electrical watts to acoustic watts is quite low, typically less than 1%. Fortunately, amplifier watts are easy to come by and our hearing is exquisitely sensitive; in practice one acoustic watt goes a very long way.

Include the virtue of very moderate cost, and noting that general purpose loudspeaker drivers may be mass produced like light bulbs, and it is these fundamental strengths that make the moving-coil principle so very effective, and so very popular. Over 99% of all the loudspeakers ever made have been moving-coil direct radiators. And the operating principle may be used over a very wide range of applications, from low-power speech reproducers of just 2.5 octaves bandwidth and a modest 75?dB maximum sound pressure output, built on a frame just 20?mm, to low-frequency capable monsters of 600?mm, capable of generating 20?Hz sound waves at body shattering 110?dB pressure levels, still more if acoustically loaded by a horn and/or aided by a local boundary.

Humans have sensitive hearing and even a low-efficiency drive unit, at typically 0.5% for the conversion of electrical to acoustic power, may be more than loud enough for many purposes. Indeed, the vast majority of domestic direct radiating speakers, including hi fi designs, are of similarly low efficiency. An electrical watt converted to sound level by this means will result in an average of 86?dB spl at 1?m, or about 80?dB in-room for a stereo pair, and in practice this is satisfactorily loud. For comparison, normal speech at 2?meters is about 70 to 73 dBA while shouting might raise 80?dB. Orchestral crescendos might raise 100?dB in room.

1.1.2 Loudspeaker Types and Technologies

Most moving-coil cone...