Although sound generation plays a major role in our everyday live, the underlying principle mainly used in todays loudspeakers has not changed since its invention in 1924. Advanced Digital Sound Reconstruction (ADSR) moves away from the classical idea, where only one driving membrane is actuated according to an audio signal and focusses on a pump-like actuator. Replacing the classical membrane with a pump enables us to modulate the airflow of the pump according to the audio signal, which generates a much larger sound pressure especially at low frequencies. In terms of a frequency response, this means that we can achieve a completely flat response regarding the sound pressure, whereas the classical analog mode is described by a gain of 20 dB per decade with increasing frequency. Since the underlying principle also relies on small structures, it is perfectly suited for the new emerging technology called Micro-Electro-Mechanical-Systems (MEMS). Utilizing a MEMS-transducer with ADSR aims to deliver an enhanced experience regarding low- to mid-frequency sound while also decreasing the energy consumption and the overall form factor due to its MEMS basis.
In our project we mainly focus on investigating this new technology of sound generation in terms of its potential as well as the design of a suitable transducer. This includes simulations utilizing the finite element method (FEM) regarding proposed transducer designs in terms of acoustic radiation with the open source software openCFS. Apart from acoustic simulations linearized flow equations are used to determine the effects of viscous damping in small structures. A coupling strategy allows us to couple the mechanical simulation of the actuator to the surrounding air, thus enabling us to fully simulate the transducer from the excitation to the final acoustic output.
Sketch of a MEMS-transducer (right) with a close up of the simulated region (left).
Another focus of the investigation of ADSR is the excitation itself. Here we focus on deriving excitation signals which optimize the sound generation by reducing the total harmonic distortion (THD). Furthermore, both macroscopic and hybrid experimental setups have been developed to prove the general concept of ADSR as well as to verify simulation-based predictions.