FEA software provides sound advice

Dean Palmer reports on how the latest design analysis techniques can be used to develop an innovative speaker system The design of a loudspeaker has undergone few changes in its long history. One of the earliest records of the use of a moving coil to power a diaphragm to generate sound was recorded in Alexander Bell's patent application for the first telephone. But the fundamentals of the design have changed little since. But finite element analysis and simulation (FEA) software is helping design engineers to better understand the mechanical behaviour that is required from the speakers and so enable more complex designs to be developed. Technology company NXT has been designing new loudspeaker technology for the past 10 years and has now developed a new class of loudspeaker known as the distributed mode loudspeaker (DML) (licensed technology from SurfaceSound and SoundVu technologies). Previous speaker technology used a driven cone to deliver sound. This restricted the quality of the sound that could be output from a single speaker for a given frequency. A large cone will produce a better low frequency signal, while a small cone will produce a better high frequency signal. This is the reason why most speaker systems will use a combination of three, different sized speakers; the 'woofer' (low frequency), the 'mid-range' and the 'tweeter' (high frequency). The diaphragm for each of these speakers is effectively rigid. This is the fundamental difference between the cone speakers and those that use the licensed DML technology. Like the conventional speaker, the panel of the DML speaker is mechanically-driven, but there could be multiple driving points and the panel is designed to undergo particular modes of free vibrations. The acoustic properties of the speaker are dramatically altered by allowing the panel to vibrate. The diaphragm dimensions no longer control the quality of the output for a given frequency. High and low frequency output can be generated from the same panel. It is important to have a thorough understanding of the mechanical and acoustical properties of the speaker structures to develop the vibrating panels, especially for those used in the DML speakers. NXT used general purpose Abaqus FEA software from Abaqus UK to help it do this. Both the speaker and the surrounding acoustic medium (the air) can be considered to be light. Therefore, a coupled approach was taken using Abaqus where the structural vibrations are influenced by the added mass effects of the surrounding acoustic medium. Using this numerical approach, it was possible to calculate the sound field generated by the panel and the acoustic field around a 'virtual listener'. By driving the panel at two points, it was possible to reproduce the 'Virtual Surround Sound' effects that can be created from a two output conventional speaker system. This effect gives the illusion that the source of the sound is at a point away from the speakers. The inputs to the two driving points are processed in such a way that the sound field around the left and right ears tricks the brain into believing that the sound is from a different source point. This effect is reproduced very well by the DML speaker system and can be predicted using Abaqus software. DML speaker technology is very attractive. The speakers can be built into everyday surfaces such as monitor screens, wall and ceiling panels, automotive trim and mobile phone screens, and make compact, very stylish designs. Pointers * NXT used FEA software to better understand the mechanical behaviour of the speaker structures and therefore develop more complex, stylish designs * Using a numerical approach, Abaqus was used to calculate the sound field generated by the speaker panel and the acoustic field around a 'virtual listener' * The software also helped predict the 'Virtual Surround Sound' effect from the speakers