Unfortunately, the choice of materials from which the plinth is made is hotly debated on many forums. In recent times, massive plinths consisting of layers of different woods or plywood were the flavour of the month. This has been replaced with stone, principally slate, but including soapstone, granite and marble.
Which are the best materials, and why? Can other materials be used? What is the role of the plinth? Can we understand its functions, and is it possible to use materials better suited to our needs? These are the questions I've set out to answer.
What is the role of a plinth for a record deck? Well, principally a plinth is the support for the deck. In the early days of music reproduction, the 'record player', amplifier and loudspeakers were all placed in the same piece of furniture. Nice to look at, but not great to listen to; one of the problems was feedback caused by vibrations from the loudspeakers being picked up by the record deck. Once the two (or three or more parts) were separated, vibrational feedback was greatly reduced.
But vibration is still a problem with record decks. The listening room is a source of vibrations, both from the loudspeakers (aerial vibrations) and from the floor (seismic vibrations). Even if these are effectively dealt with, the record deck itself can be a source of vibrations. So a secondary role of a plinth is to reduce any vibrations coming from the deck to a low level, so that it isn't picked up by the stylus/cartridge, and does not interfere with the reproduction of the recorded waveform.
A plinth may be a simple wooden box, though this is usually quite resonant.
or a complex hitech design, like this one, a rendition of my G99 turntable,
And this is what my G99 looks like at the moment:
©Cat's squirrel 2010
When designing a plinth, the choice of materials is very important. Any vibrations from the deck itself need to be reduced by the plinth, as well as aerial and seismic intrusions. Not many materials will reduce these vibrations, their damping factor needs to assessed, and should be above 0.2 to have any real effect.
Any material chosen, if thought of as a panel, will have certain characteristics. There are at least three main frequency regions where vibrations are controlled (or not) by different means. The lowest frequency region is controlled by the stiffness of the material. This is where the panel deflects as a whole plane. The stiffer the material, the less the deflection (vibrations), but the lower the critical frequency will be. Then there is a region where the fundamental resonance is found. This resonance frequency will depend on the dimensions of the panel, and the speed of sound of a wave, which is dependent on the material's Young's modulus, density and Poisson ratio. The fundamental resonance amplitude will usually be the highest of any mode, the sharpness (Q or quality factor) depending on the damping factor. The next region is controlled by the mass of the panel, and the higher the mass the better, but the resonance peaks will be affected by the damping factor, and the thickness will have a dramatic affect on the critical frequency. This is the frequency at which the panel will be transparent to sounds, again affected by the damping factor. Above the critical frequency is a region controlled by damping.
So it is vitally important to have as much damping as possible, for the material/s to be stiff, but not too stiff, and have sufficient mass, but not be too thick. Balance all that out and you will be well on your way to designing a decent plinth.
But which materials are suitable, and what thicknesses are best? Fortunately, I've been working on some very involved calculations which seem to answer both questions. By knowing the materials properties like Young's modulus, Poisson ratio, damping factor, density and the dimensions of the panel, its possible to work out the sound losses at any frequency. A graph can be plotted, of frequency against losses, which gives the response of the panel, like this:
What you see above is the sound losses from a panel, the losses are due to heat generated from friction and from sound radiation into the ambient surroundings. The graph above shows the losses from a panel of Delrin, 18mm thick. At 20 Hz, the loss is about 18dB, which decreases until about 100 Hz, the fundamental resonance frequency. Up to this frequency, the panel losses are controlled by the stiffness of the panel Around 100Hz, the losses are only about 10dB, not really good enough. The peak height is controlled by the intrinsic damping of the material of Delrin, which has a low damping factor. Above this is the mass controlled region, where losses continue to increase until a very deep valley is reached. This is the critical frequency, and the depth is controlled by damping (or not, in this case). The panel will be completely transparent to sound, with no losses in this region, so expect a sonic signature which will appear to amplify all sounds around this critical frequency. Above this region, losses increase at a faster rate, until the dilatational frequency, but it is so high in this case, it will have little affect.
The graph above is for polyester cast resin, It has excellent losses from 20Hz upwards (at about 35dB) a low fundamental frequency peak, due to the very high damping factor, and even a critical frequency that is 20dB down. There should be no 'added sparkle' from the peak due to the dilatational frequency, as it is well above the normal range, about 12 kHz. For a single material, this is about as good as it gets.
To enable comparison of materials and thickness, I have devised a goodness factor, which integrates under the curve from 20 Hz to 1000 Hz, by summing all the data points. The equations I derived were used to determine the optimum thickness for each material listed below. It appears that thicker is not better, as the thicker the plinth, the lower the critical frequency and the higher the fundamental resonance frequency. The size of each panel was set at 400mm x 350mm, chosen as a typical value for plinths:
|thickness||factor||frequency||frequency||frequency||loss at 20Hz||loss at 1kHz|
|mm||(20Hz – 1kHz)||Hz||Hz||dB||dB||dB|
|polyester resin||24||2541||268||1581||33 ||34||41|
Note, all data must be accepted as provisional until the calculations have been validated. Last updated 20th Oct 2k10
Also note, values may change a little as calculations are refined.
From the above mentioned 'goodness factor', it may come as no surprise that the BBC had devised something similar to describe panels used for loudspeaker construction. Their equation for their 'figure of merit' is:
SQRT(Young's modulus x density)/Q, which, at resonance equates to SQRT(YM x density) x damping factor
the same three factors which are necessary to consider for plinth materials, or any panels for audio work.
Of course, the information in the table above can be used for panels other than for making plinths, it is just as applicable for loudspeaker panels and support shelves, although it should be pointed out that the figures shown above are for a 'simply supported panel'. A box (as in a loudspeaker) may be more difficult to model.
But is a thick plinth better than a thin plinth. Surprisingly, the answer is no. Take a look at this short video, it illustrates what effect thickness has on frequency response. It is, perhaps, counter intuitive, it doesn't make sense, but nonetheless, the optimum thickness of plywood for a panel this size is just 6mm.
I have now worked out equations for two layer panels and three layer sandwich composites, (assumed to have top and bottom layers of the same material and thickness) and five layer composites. The latter can be used to get a better feeling for panel responses by modeling the adhesive layer in three layer laminates. It really can make a difference, and again, thick glue layers are not necessarily better! I need to validate the spreadsheets, but I have decided, if all looks OK, to post the findings on a separate website.
Watch this space for details, (it will be on the side panel).
A note on c.l.d. - constrained layer damping:
Although this has a place in audio builds, it is a term that is usually misapplied when talking about plinths. True cld is applied to a vibrating panel, to reduce the vibration amplitudes. It comprises a layer of viscoelastic material (ve) stuck to the panel which is vibrating, and the ve layer is covered with a stiff layer, similar or identical to the vibrating panel, but often thinner than it. Thus a typical application may be an aluminium/ve/aluminium composite. However, the ve layer moves sideways, and looses energy by shearing. More complex arrangements might have several layers, or strips or stand offs, but the principle is always the same, shearing forces are used to lose energy (actually it is converted to heat energy).
What it is not. Usually, the term cld is applied to a multilayer of wood, either the same type, as in plywood, or different types, as in ply/mdf. These are not cld, as there is no viscoelastic layer (the layer which converts the energy). It is called a glulam, a contraction of the term 'glued laminate', used in buildings as beams and other constructions.
A note about resonance and vibrations:
These words are often used interchangeably in audio, although they can be very different things. We all know what a vibration is, something which is vibrating has all, or most, of its structure moving in a way we call simple harmonic motion, although the motion may be anything but simple. A resonance is also a vibration, but one where the input energy is causing the maximum amount of vibration. When a panel is hit, it vibrates, and quickly settles to a single (or several related) tone/s. So what is the difference? The resonance frequency of a structure is determined by the material it is made of, and its dimensions. So there will be a set of harmonically related resonance frequencies. However, the same structure can be made to vibrate at almost any frequency, except their amplitudes will be far less than for resonance frequencies.
So a panel used as a plinth will have a set of discrete resonance frequencies, which will only be a problem if excited. If a motor is connected to the plinth, it will vibrate at a certain frequency, and it will start the plinth vibrating at that frequency, also. This is called forced vibration, and is very different from resonance. Forced vibrations can be any (sensible) frequencies, whereas resonance frequencies are very specific to the material and dimensions.
And relating to our plinth, airborne vibrations will cause the plinth to resonate, whereas the motor (and seismic intrusions) will cause the plinth to vibrate. But, the motor vibrations will build in the plinth until the intrinsic damping can bring it to a steady level, where the energy lost equals the energy input. Plinth materials with low intrinsic damping will have much higher vibrational amplitudes (caused by continuous forced vibrations) than those materials with higher intrinsic damping.