What is Constrained Layer Damping?
Constrained Layer Damping (CLD) came to mind when we decided to design the ST 15 by adapting the ST12i to fit the K-Series and Elan ranges. We changed the dimensions and the top from separate difficult-to-assemble strips to a one-piece plate. However, with the first samples we noticed immediately that the sound deteriorated, although the basic configuration and mass were largely similar to the original.
The main problem was traced to complex modal patterns (resonances) in the new top plate and the base plate, followed by lower-level resonances in the steel support tubes. The first step was to replace the all-over top plate with one with a rectangular hole in the middle, which reduced the complexity and increased the frequency of modes in the top plate. But now we were on a mission to improve the sound further and we felt this alone was not enough.
The cure was to damp all the metal parts. Many methods were tried. The obvious one was simply to add damping material, such as the thin bitumen or polymer sheets used in car manufacture to damp thin metal panels. This did very little to tame resonances. The 4mm steel , used for optimum strength and stiffness in the top and bottom plates, is heavier and stiffer than automobile steel. The extra stiffness and mass store more energy and tend to increase the Q of resonances, so that they ‘ring’ more strongly and for longer. We would need very thick, heavy (and ugly) conventional damping panels to have any effect at all, and they would add considerable mass. (As more damping mass is added, even more damping is required, due to the increased energy potentially stored). Clearly, conventional surface damping was not the best way to tackle the problem.
Constrained Layer Damping
Constrained Layer Damping works entirely differently to simple surface damping. In CLD, a relatively thin, low-mass damping layer is trapped, or constrained between the surface to be damped and a stiff counter layer, so that when the surface vibrates, the mechanical energy is ‘trapped’ between the two and dissipated in the constrained layer as heat. So, CLD is less bulky and more effective than surface damping, while minimising the additional mass. To work well, the theory states that counter panel must be sufficiently stiff, and the damping layer’s mechanical impedance matched to the panel and counter panel for best energy transfer and absorption.
Conventional surface damping
Constrained layer damping
So, we embarked on a development program to try various constrained damping layers and counter panels and find the ones that worked the best. The differences are measurable using accelerometers, and while we did many such measurements (a selection are shown below), they do not necessarily relate directly to the sound. So we mostly listened to the resonances in the stand and the affects they had on the sound of the speakers. We tried damping materials designed specifically for damping and ones which were not. Surprisingly, we found one that worked far better than those designed for this purpose and this was also less expensive. Also, having experimented with aluminium and steel for the counter-panel material, we chose steel as the most cost effective. The counter panel (and constrained damping layer) are very thin and, being fixed underneath the top and bottom plates, are only visible if the stand is inverted.
The vertical support tubes were more problematical. It was not practical to insert tubes and such a thin damping layer. Fortunately the resonances in the tubes were much lower in magnitude than in the plates, at higher frequencies, and required only modest damping, which we achieved by injecting expanding foam into the tubes after the ends had been welded, painted and heat cured. The biggest reduction on resonance was due to CLD, with the tube damping adding the ‘cherry on the cake’.
Below are shown examples of results from accelerometer tests on an un-damped and damped ST15 stand (CLD to the top and bottom plates and foam in the vertical tubes). All measurements were made with an Epos Epic 1 loudspeaker, mounted on four small pieces of Blu Tak.
Effect of CLD on the bottom plate near the front left hand column: green undamped, red, with CLD. Note 8 to 10dB reduction in acceleration over a wide band between 600Hz and 2 kHz. There is an increase in acceleration at 280Hz, but this is only over a narrow band and is more than offset by the broad and deep reductions across the mid range.
Acceleration in centre of speaker top panel, green without damping, red with CLD, showing approx 5dB reduction in acceleration over the band between 2-5kHz where the ear is most sensitive.
Acceleration on bottom plate in centre: green without damping, red with CLD, approx. 5dB reduction 1 kHz-2 kHz.
Note that -3dB represents a 30% reduction in magnitude, (or half the power), and -8dB is about 40% reduction in magnitude, (or less than 1/6 power), over quite a wide band of frequencies, so the reduction in vibrational energy, especially in the critical mid band is large and very real, affecting not just the stand, but the speaker cabinet. So, it should be no surprise that the effect on the speaker’s sound was dramatic.
CLD proved to be highly successful. With the application of CLD the music was now more open, freer and less cluttered. There was no slowing of bass due the increased damping– in fact the bass opened out, becoming faster and more fluid, while the whole sound stage became deeper and wider. In fact the speakers sounded less ‘boxy’ and like much more expensive models. That was when we knew we had achieved the best result, given acceptable aesthetics and practical manufacturability, at a sensible price. The cost of the stand has increased, and the end selling price with it, but we feel that the improvements are well worth it.