How Should Speakers be Stabilised?

[sdolezalek]

This topic never ceases to amuse me.  On a website dedicated to so many arguments between the importance of measurements and a view that my ears are all that matter, this is a topic where measurements should dominate, but the predominate reason they don't is that people complain about the math being too hard to understand.  Instead they retreat to the comfort of what sounds like a conservative vs. liberal argument: conservatives want rigidity and coupling, to as much mass as possible, liberals want to maximize flexibility in all directions; neither side having much more than a very casual understanding of why they believe what they believe.

 

In addition, the same answer is applied to vastly different components -- in a speaker the diaphragm moves (vibrates)  and it must to produce the sounds we want to hear.   In an amplifier or other piece of solid state equipment, are you trying to shield the equipment from room vibrations or shield the room from equipment vibrations? 

 

Some obvious questions:

 

a)   Just how much mass do you need to couple to an 18 inch subwoofer to keep its enclosure from making any movement? Is a cabinet enough? Do i need the whole room?

b)   If I optimally float my subwoofer does that mean that the cabinet moves in exact opposite response to the speaker cone, thereby resulting in no sound?  If the cabinet moves differently than the speaker cone, i.e. half as much, or only in certain frequencies, or only in delayed fashion, what happens?

c)   If, in theory, I'm trying to convert vibration energy to thermal energy what are the rules of physics I'm applying?

d)   If I couple my speakers to the room, should I float all my equipment so the vibrations from the room don't flow into the equipment?

e)   If I have coupled my equipment with spikes and weights, how much do I need to worry about room, speaker or ground-based vibrations?

f)   If I'm using a roller-block system, what frequency should the rebound motion of the roller be tuned to, does that change between speakers and electronics?

g)   How does absorption (i.e. pieces of vibration absorbing foam), differ from large mass coupling, differ from flotation?

h)   if I have accomplished my objective, what difference would I see in a sound plot of my system? Sharper transients? Less reverb? 

 

I have zero doubt that doing the right thing here can make a big audible difference, but it is amazing that in this age of big data analytics, we have so little concurrence on a topic like this. 

[sdolezalek]

On 12/28/2018 at 7:40 PM, look&listen said:

WTF? Much can be learned from writings of Barry Diamont & John Swenson on theory of operation of roller isolation & more. Much good discussion here if search older posts! I soon look at recommended Sistrum documentation.

 

Have mental model of roller isolation as mass of supported device tending to stay at rest (like physics law) while rest of world buzzes around under it. Issues of retention, materials, friction, mechanical impedance, hardness, concentricity. e&. there too, as implementation details, but takeaway is physics to rescue  

 

So maybe better for you to dig deeper before announce a pox on both houses?

I have very carefully read all of the Barry Diament threads on this topic (and virtually all of the other vibration threads on Computer Audiophile).  Barry's methods make a huge amount of sense in terms of isolating equipment from outside vibrations.  His roller bearings (if built correctly) should isolate equipment from horizontal vibrations quite effectively, but can't do much in a vertical direction (which is why he also uses inner tubes (or air bearings) to create some level of up and down isolation. 

 

But think about what his roller bearings do for his Magnepans.  Every spring is tuned to be more or less compliant at different frequencies, in other words at some frequencies the spring acts as a highly effective isolator and at other frequencies it behaves just like a metal spike.  I believe that Barry has worked to make sure the natural frequency of his roller bearings is below the audible range of his Magnepans.  If that is true, then they should act as spikes (stiffeners) for all audible frequencies, while effectively isolating any nonaudible frequencies that might still affect the frame of the speaker.  Separately, if you look at the designs of many other "roller bearing" applications, what you see are fairly small dishes in which the sides roll up quickly -- which means that the amount of isolation is highly limited, probably to far less that the excursion limits of a 20Hz woofer.  What does that mean as a low frequency driver plays softer and louder notes? Are the softer notes fully isolated, whereas in the louder ones the excursion limits are reached?  

 

As to the Sistrum system, I loved the following review: "Sistrum Rack Systems allow the speaker generated resonances and the associated airborne resonances to happen. If it vibrates; let it. This is a concept that we just don't hear about. Absorb, absorb. Isolate, isolate! Well, their speaker support system is a complex vibrating conduit, composed of a secret (patent pending) metallurgy formula, and added geometrical design that directs, very expeditiously, resonances to earth's ground with no back-feeding effects."  Sure enough, if I can just make the resonances go away by "sending them to ground" then I don't need to worry about either isolation or absorption...  You just need the right metals and geometries...

 

I'm not so much trying to announce a pox on either absorption or isolation (or for that matter stiffening), I just think that in this time of computing, analytics, big data, etc. we can do better than most of the explanations I see as to how or why a certain technology works, or what it is trying to protect us from.