Lowering Mechanical Noise Floor in Speakers - page 4

Lastly, I varied the panel material's Young's modulus, a mechanical parameter that effectively characterizes a panel's stiffness. As expected, we see that IL response variations happen primarily in the lowest decades of the IL response spectrum; IL increases as E increases. Note too that with increased stiffness we see the first, lowest frequency IL minimum shift upward in frequency.

To complete the simulation run, I applied various simulated IL response spectra to actual db-SPL driver response curves, which resulted in predicted panel acoustic response, when driven by the acoustic pressure waves generated by the driver enclosed within the simulated polymer composite cabinet. The simulations were then run, producing predicted results of dB-SPL measurements, taken at 1 meter, outside the cabinet.

Baseline

The top curve is the dB-SPL response of the driver, measured at one meter. The bottom curve is that of the predicted response as measured at a distance of 1 meter outside the simulated cabinet that totally encloses the driver.

Keep in mind these results are for an enclosure that contains no sort of sound-absorbing filling or liner; thus the thermodynamic condition obtaining within the enclosure is adiabatic. As well, the simulation does not take into account panel acoustic output resulting as a consequence of mechanical vibrations originating at the driver's frame.

At any rate, once having established a performance baseline I then ran a further set of simulations, each with the purpose of exploring the effects of varying mechanical properties of the cabinet panel's material. (Keep in mind that in each of the following simulations the top curve hovering around 100 dB Spl is that of the driver when not enclosed by the cabinet.)

Density Variations

My first simulation series held all properties constant, except for density. Varying density produced substantial differences in the amount of insertion loss and in turn substantial affects on panel acoustic output. The resulting curves show that more dense the material, the higher the insertion loss, resulting in a decreased panel dB SPL levels.

Next, I re-ran the simulation, this time holding all parameters constant, save stiffness. Increasing stiffness did

have an effect on IL, but it took quite a substantial increase in panel stiffness before significant changes

in IL and thus panel acoustic output became evident. We can see from the graph that IL did increase as stiffness increased, and at the lower end of the response spectrum, just as predicted by theory.

Stiffness Variations

Last, I ran a series of simulations holding all parameters constant, save thickness. As we can see in the accompanying graph, increasing panel thickness increases IL.

Conclusion

In this article we have seen that panel resonances, either structural or airborne in origin, can have effects, usually deleterious in nature, on system response. We have also seen that the designer has at his or her disposal a effective means by which these resonances can be damped, dispersed or otherwise shifted out of the frequency range of interest. I think it also fair to say that the use of such modeling techniques as FEA can be an invaluable, efficient tool where it comes to exploring various design solutions when seeking to lower the mechanical noise floor of the system.