Abstract:

In the constant displacement region, several displacement-dependent quantities control the behavior of electro-dynamic transducers. This research uses a tensile/compression loadframe to measure the displacement-dependent stiffness, B1-Product, and mechanical losses of various driver components. The individual data sets are curve-fitted by power series expansions. The power series coefficients are used to predict the relative low-frequency total harmonic distortion (THD). Then, the measured components are built into loudspeakers, and the predictions are compared to actual THD measurements. Experimentally, this method is very sensitive. Additionally, the effective stiffness of a transducer's suspension is merely the sum of the power series coefficients representing the spider's and surround's stiffnesses. By comparing various spiders and cones, the effective stiffness of a suspension can be linearized, thereby reducing low-frequency THD. Finite-element analysis (FEA) simulation of nonlinear suspension behavior can facilitate such a linearization process. Lastly, power series coefficients can determine the relative variability of components (e.g., which of two spider designs is more variable). OEM manufacturers of transducers must maintain statistical process control of Q[inf Tot], resonance, and THD. Q[inf Tot], resonance, and THD are all dependent on displacement-dependent quantities. Therefore, reducing component variability is helpful in achieving an acceptable c[inf pk] for Q[inf Tot], resonance, and THD.