Speaker Driver: Comparison of Options
Speaker driver choice is a very important consideration, since the transducers themselves are of course the most fundamental part of the speaker. Regardless of other factors, one can never expect inferior drivers (and hence the system as a whole) to perform well.
There are two main options when choosing drivers; electrostatic or conventional voice-coil designs. Although many seem under the impression that electrostatic loudspeakers are a modern invention this is not the case; Janszen was granted the first U.S. patent for such a device in 1953[1]. Considering the relatively small market penetration of electrostatic transducers and the fact that they tend to appear largely in high-end designs, one might be led to assume that electrostatic panels are superior to conventional drivers. This however is only partially true.
One advantage of electrostatic panels is that full-range designs are possible, eliminating the need for crossovers and hence the associated problems with frequency and phase response in the crossover band. Another advantage is that the electrostatic panel is generally very light and hence offers excellent transient response, whilst also offering very good directionality and imaging. The latter may also be seen as disadvantage, since it effectively makes the ideal listening position rather narrow.
In terms of disadvantages, the chief problem with electrostatic designs is a difficulty in reproducing bass frequencies at high SPLs. Generally the panel excursion is small, which makes it hard for electrostatic transducers to move the required volume of air at low frequencies. Furthermore, since electrostatic transducers are not meant for use with an enclosure, phase cancellation is an issue, again resulting in reduced bass performance. Audiostatic, a company that manufactures audiophile full-range electrostatic speakers, admit of their own devices with regard to bass that “Obviously because of the limited membrane excursion they won’t produce ear shattering levels at that frequency”[2].
As a result of the aforementioned bass performance, many high-end electrostatic speakers are in fact hybrids, using voice-coil woofers for low frequencies with electrostatic panels covering the mid and high range. One example is the Martin Logan Summit[3], which whilst described as “our most advanced and sophisticated full-range loudspeaker” nevertheless makes use of two 10” woofers for low-end reproduction. Of course in this situation a crossover is still required, so the advantage of the possibility of a full-range design is often nullified in practice. Still, electrostatics may prove very attractive as high quality mid to high frequency drivers, although they are certainly not cheap.
In choosing conventional voice-coil drivers, there are many factors to consider. In terms of quality, it is certainly true that one does indeed get what one pays for. Whilst high quality manufacturers such as SEAS[4] are happy to provide detailed frequency response plots and Thiele-Small parameters for their transducers, many cheaper manufacturers are less transparent about their devices.
One common trick to beware of, often used by less scrupulous manufacturers, is the quoting of a recommended frequency range without stating the variation in output (in dB) across this range. A recommended operating range without any indication of the actual performance within the frequency band is virtually meaningless. Many assume a ±3dB range is implied when reading such data; it is unwise to make such assumptions.
Furthermore, even if frequency response across a range is qualified with the variation in output in dB, this is still not ideal. Obviously one desires that any variation in output magnitude will be a smooth variation; one still has no idea of how “lumpy” the response might be. For these reasons it is best to choose drivers that are accompanied by frequency plots, since this gives a far more accurate representation of true performance.
Another important consideration in choosing a driver is the application for which it is intended. For example, a woofer with a high maximum cone excursion and low Fs may perform very well in a large sealed cabinet but be totally unsuited to a ported implementation (Dickason, 2000). One can make use of the quoted Thiele-Small parameters to ascertain whether the driver is suitable for its intended purpose.
Construction materials also give an indication of how the driver may sound. In terms of woofer and midrange drivers, for example, an aluminium cone may indicate greater bass precision than an otherwise equivalent transducer with a paper cone; softer cones are associated with greater distortion than their stiffer counterparts. However, as Larsen (2003) notes “cone break-up behaviour and frequency response was shown to be strongly dependant on the Geometrical Stiffness of the Cone”. Hence the geometry of the design may be more important than the material used.
Diameter of the driver is also a hugely important factor for woofers, although of minor importance for tweeters. To reproduce bass frequencies at good SPLs, a large volume of air must be moved by the driver. To this end, there is absolutely no way a 6” driver can compete with a 12” driver of similar quality in terms of bass extension; it is simply not physically possible.
Power handling is another consideration that must be given thought when choosing a driver; the peak short-term power dissipated by a transducer can easily be double its long-term rating. Naturally for the best performance it is desirable to ensure that the driver is not operating too close to its quoted limits. One should think carefully about how hard the driver is likely to be driven and ensure its power handling is adequate; overdriving a unit at best will result in distortion and at worst may cause irreversible damage. In many cases users overdrive and damage units in an attempt to achieve a higher SPL, particularly in the bass region. If the system requirements are adequately specified and designed for, this should not happen.
For the high-budget client, the best solution will either be high-quality voice-coil drivers carefully selected to complement each other, or a hybrid electrostatic implementation. It is difficult to recommend a fully electrostatic solution due to the associated problems with low frequency performance, although for some clients this may be acceptable.
For the low-budget client, standard voice-coil drivers are the only solution. The quality of the drivers used will largely be influenced by pricing; one should carefully consider all factors and attempt to find the best solution within budget. Datasheets should be closely scrutinised to identify the strengths and weaknesses of each option before a solution is chosen.
In conclusion, notwithstanding the electrostatic debate, driver choice is largely influenced by price and performance. In general, the better specified the driver, the more expensive it is likely to be. If working with a high budget, one is likely to simply choose the best specified drivers. Conversely, with a limited amount of capital, one must make the best compromise that can be reached within budget.
Sources
Larsen, Peter. (2003). Geometrical Stiffness of Loudspeaker Cones, Loudsoft.
Borwick, John. (2001). Loudspeaker and Headphone Handbook, Focal Press.
Dickason, V. (1995). The Loudspeaker Design Cookbook, Audio Amateur Publications.
Rossing, T. (1990). The Science of Sound, Addison-Wesley.
1
