Dynamic mass modification by electric circuits

Abstract



There are two concentrations in this project. One is to explore the possibility to construct negative acoustic impedance by electronic techniques, and the other is to see whether such method can be utilized to build effective sound absorber using electromagnetic actuator (here we adopt the moving-coil loudspeaker as sample) with a shunt circuit.

Our study begins with analytical analysis, and the result shows that it is impossible to gain independent control of basic acoustic impedance components (mass, stiffness and damping) by simple circuits. Two alternative designs are put forward as a compromise. One is the series circuit with NIC to simulate the negative acoustic impedance, and another is the series-parallel circuit with NIC. Theoretical prediction shows that we can indeed obtain broadband negative mass and local negative stiffness by these two types of circuits, and that we can achieve broadband noise control with simple electronic shunt circuits despite fact that completely independent control over each parameter is not possible. We argue that these conclusions represent significant technological and economic advantage worthy of further development.

All analytical results are validated by experiments with satisfactory agreement. The sample loudspeaker with shunt circuit is tested with acoustic impedance tube. The rig consists of a DC powered op-amp circuit and a loudspeaker. An efficient Matlab code controls the excitation sound generation and data acquisition with AD/DA cards.

Two typical and most interesting results are summarized here. In the first, a series type circuit with NIC is used to construct negative equivalent mass and local (banded in frequency domain) negative stiffness. We experimentally demonstrated that it is rather easy to reduce original mass of the loudspeaker to half of its original value and it could be reduced to almost zero. This is evidenced by a very flat sound absorption coefficient curve from 100 Hz to 1000 Hz.

The second circuit is a series-parallel circuit. It’s an improved design from the first type. This type of circuit can, to a certain extent, decouple the stiffness and mass controls. The results show that we can reduce mass globally (in the frequency domain) and stiffness at low frequencies. The original mass of the sample loudspeaker is almost eliminated and the stiffness at low frequencies is reduced too. In terms of the spectrum of sound absorption coefficient, it manifests itself through a broadband absorption with prominent improvement in the low frequency region.

Finally, potential applications for our designs are discussed. A tunable low frequency resonance absorber is designed. Prediction results point out that, by choosing the right parameters of the circuit, we can achieve 100% absorption at any given low frequency. Thin absorber is another potential application. With the same dimension, the performance of a thin absorber is much better than that of the standard sound absorption construction. A 90% noise absorption from 300Hz-600Hz and 50% absorption from 250Hz-1000Hz is achieved by our new design. The dimension can be further reduced in theory. Finally, a broad-band absorber with 50% or more absorption over 80Hz-1000Hz is demonstrated.