Basis established for new generation of loudspeakers

June 04, 2021

Monday 31 May 21Mathematical model makes it possible to design future loudspeakers with a shape and in a material which significantly improve sound quality and which design engineers would hardly have come up with on their own. A recently completed research project at DTU has developed a new mathematical model for pinpointing changes in the design of loudspeakers which result in significant sound reproduction improvements. Based on algorithms for how the sound reproduction quality should be, the computer calculates the optimal choice of shapes and materials. The researchers behind the project have no doubt that the new optimization model will provide a basis for major changes in the future design of loudspeakers. “We’ve no doubt that the model works, because all the elements that form part of its calculation basis come from knowledge previously tested in practice.

Monday 31 May 21

Mathematical model makes it possible to design future loudspeakers with a shape and in a material which significantly improve sound quality and which design engineers would hardly have come up with on their own.

A recently completed research project at DTU has developed a new mathematical model for pinpointing changes in the design of loudspeakers which result in significant sound reproduction improvements. Based on algorithms for how the sound reproduction quality should be, the computer calculates the optimal choice of shapes and materials. The point of departure has been that the sound must be reproduced so that it is as close as possible to the sound originally recorded by the artist. In addition, the sound must be dispersed well in the room, and all frequencies must be reproduced in the same way to avoid too much bass or treble. 

“For example, we’ve used the model to calculate how an acoustic lens that ensures dispersion of sound in a compact smart loudspeaker is to be designed to give the best result. Here, the computer’s calculations showed that the optimal lens geometry is more curved than conventionally; it’s almost s-shaped. This gives an even frequency response and good sound dispersion,” says Daniel Gert Nielsen, who is behind the new model. 

The model has been developed in collaboration with researchers at DTU Electrical Engineering and DTU Mechanical Engineering, which are world leaders in the combination of acoustics and the impact of vibrations on mechanical components. 

Choice of materials and new production methods
The model has also been used to look at the choice of materials for some loudspeaker components. A base of a hundred materials was fed into the model, which subsequently made suggestions to the material layout that differ from the usual materials. 

“ In many cases, the choice of materials in the membranes of existing loudspeakers is based on the relationship between the weight and the stiffness of the material and its damping properties. The model suggested that the material properties of the membrane should not be identical all the way through. In some places, it should be heavy and soft and in other places light and stiff. This variation allows for the best reproduction of sound,” says Daniel.

The researchers behind the project have no doubt that the new optimization model will provide a basis for major changes in the future design of loudspeakers. This is an area in which the limits have been reached for what can be achieved using the more conventional design methods, and where, for example, computer simulations and new production technologies can open up for even better audio experiences in a future generation of loudspeakers. 

“The development of 3D printing technology and the use of advanced composite materials mean that, in future, you can produce a loudspeaker component—for example a membrane—in which the material varies as suggested by our computer model,” says Daniel. 

Students are invited to continue the work

Daniel is pleased that he chose to spend a couple of years doing research after completing his engineering studies. Now he hopes that the foundation he has created with the development of the computer model will be picked up by new students who, for example, are interested in building compact smart loudspeakers based on the model’s calculations. 

“We’ve no doubt that the model works, because all the elements that form part of its calculation basis come from knowledge previously tested in practice. But it could be interesting to conduct practical experiments if some of our students feel like trying their hand at the task. Daniel has fortunately promised that he will participate as a supervisor from his new position in the corporate sector,” says Associate Professor Finn T. Agerkvist, DTU Electrical Engineering, who has been Daniel’s supervisor.

After having concluded the research project, Daniel contacted the headset/speakerphone company Jabra to enquire whether they needed a developer with his competences. Today, Daniel is part of the company’s development team, but he would like to continue his collaboration with DTU on projects related to the loudspeaker model he has developed and to Jabra’s work in general.

The source of this news is from Technical University of Denmark