Experiment time ( or fun with rubber )

OK, Eric, I'm really trying to get this right but it never seems to surface all in the same post - are you saying that the Mass-spring-mass I referred to isn't ALSO lowered by gluing panels together? Does gluing ONLY affect the coincidence dip, which in gypsum wallboard is around 2-4 kHz? I would really appreciate it if you could clarify BOTH of these in one post - What, if any, effect will gluing have on Mass-spring-mass? Then, what, if any, effect will gluing have on Coincidence Dip? And finally, if lowering resonance below audibility helps TL at bass frequencies, why wouldn't it work to use closer spaced studs and glued panels in ONE leaf of a wall, then 2-3 different thicknesses of screwed gypsum panels on RC (over 24" centers) on the OTHER leaf? Maybe a better way would be if you could give an example of a really effective drywall assembly, then explain WHAT makes it good? I'm basically just trying to "put this to bed", or in other words, settle the confusion, in a way that is simple enough for us non-acousticians to grasp - If you can help me do this, I'll definitely owe you at least half a dozen beers :? Not to mention being just plain grateful... Steve

Hello Steve, I know I better should find some more time to explain all different phenomena in detail on a page with graphs. Lets first speak about the strongest phenomena of double leave systems: 1) MASS-SPRING-MASS RESONANCE Basically the main calculation starts from the Mass-Spring-Mass resonance, let's call it further fo. This is the most powerfull source defining the acoustic behavior. Theoretically at straight incidence: Assume a complete undamped theoretical system: fo = insulation 0 dB. In practice there is damping by the connection structure edge clamping and whatever. So fo is not 0 dB but its the lowest point. From there the TL rizes theoretically with 18 dB per octave in practice with 12 dB per octave. Starting point of this inclining line is "fo" at the level of the mass-law of both leaves together. So it's clear the lower this point the better since it also influences the subsequent higher frequencies, in fact you shift this inclining straight line to the left. Improving a double leave wall is mainly trying to get this "fo" as low as possible. Al the rest is secundary. This is defined by the combination of the 2 masses (both wall leaves) versus the stifness of the spring. You increase insulation by: a) increasing the masses (more layers) b) weakening the spring (increasing cavity width) Note that I don't speak about waves here: the cavity is much too small to ellow those long low frequency waves to develop. The cavity acts as a pressure room, which is about comparable with sound in a room far below the lowest room modes. That's why wavelengths do not matter here the air is only a spring. Here you see something fun: A drywall = exactly equal to a panel bass trap. Just flip the TL graph upside-down, substitute the TL values by absorption values and see, there you are. The Q of this dip depends on weight versus spring ratios damping and so on. This inclination of 12 dB/octave changes in 6 dB/octave when point 2 starts: 2) THE CAVITY RESONANCES Exact as in a room there are cavity resonances or modes, where the first cavity resonance is located at 1/2 wavelength. However the width of this effect causes this downward bend from 12 to 6 dB per octave to start earlier at 1/4 wavelength or even below. Then we come to the 3th phenomena: 3) THE COINCIDENCE FREQUENCY in US called CRITICAL FREQUENCY. This is where this previous described rize of 6 dB/oct bends down the a decline of the curve towards the coincidence frequency. Here I copy another message including the link you know:

Look at: http://www.domesticsoundproofing.co.uk/tloss.htm If you want the math I suggest to get a book in the library about acoustics. The coincidence freq., in the states more referred to as the critical frequency. This is the frequency where the wavelength of the incident soundwaves equal the bendingwaves in the panel itself. This causes a serious transmission loss dip. Some general things: Doubling the thickness of a panel will lower this frequency with 1 octave. It depends on material properties. E.g. this frequency positioning is better for drywall than for wood, making drywall better for TL. This frequency is angle of incidence dependent. It is about the only frequency where the radiated noise on the receiver side preserves the direction of the incident noise at the source side. Below this frequency the bending waves in the panel become shorter than the wavelength of the incident energy. This causes the radiation ratio of the panel to become lower than 100%. What happens is that the air pressure toggles between the crests and the throughs of the bending waves of the panel. As such only part of the vibration energy is transformed in noise again. In fact one could interpret this a bit as natural anti-noise (poetic licence). That's why things as drywall are such efficient insulators. Above this frequency the bending waves in the panel become larger than the wave length of the incident energy. This causes the radiation ratio to become 100%. Every bit of vibration is converted again to noise at the receiver side. The wall act a bit as a piston. In my language (literally translation) we call a wall with very low coincidence a bending-stiff wall (as e.g. a concrete or brique wall), while a wall with very high coincidence we call a bending-soft wall (as steel plate, drywall, particle board etc.). In US one refers to the latter as a flexible wall or panel. Best regards Eric

Now: If you once should find the time to play with the excel files on my site you REALLY should get more the feel for those concepts. Now a good wall is: 1 ) a lot of layers on both sides (increased mass - the exact relative effect can be calculated with my files) 2) a cavity as large as possible (weakened spring - the exact relative effect can be calculated with the same mass-spring file) 3) If using wooden studs, disconnect them with RC on ONE SIDE ONLY. If doing on both sides and making your wall theiretically more perfect THAT'S A DISADVANTAGE, since then you loose the damping of the system causing the "fo" to become TOO DEEP, making the wall WORSE. 4) try to limit the coincidence dip a bit, but DO NOT EMPHAZISE THIS TOO MUCH, since music noise is as overall insulation hardly improved by maximizing this phenomena. THe ways you can do that depends on the number of layers. Take panels with different thicknesses with each there own coincidence frequency. This will cause a damping of this dip by dismatching bendingwave lengths. It's enough that you put a 6 mm or 9 mm panel in-between, on the outside or whatever. ANY FURTHER IMPROVEMENT will NOT show in the overal TL. I even do not believe that this special screws to connect gyproc to one another rather then to wooden studs is an advantage when one side is disconnected by RC. It even can be a disadvantage: if this stiffens the panels too much by acting as a single leave causing the coincidence to lower again. Best regards Eric, This isn't complete, but to time consuming. Basically a good wall is a simple one, all the rest is magic. If I should build it for myself now: With metal studs: Direct connection to studs filling with something as 702 (not 703). Number of layers and cavity depending on insulation I need. Thickness of at least one board on both sides strongly deviating from the others e.g. a 1/4 or 3/8" panel. Both sides symetrical or not (doesn't matter that much) With wooden studs: The same but one sided with RC or comparable If using RC I should do the side without RC at a spacing 16" If I should want to have a havier wall I use staggered studs or independent frames at ca 16" OC Eric PS: PLay with the acoustic selector Excel file. You can alter the TL values from existing BORAL measurements, to imediately see what the effect is on the overal insulation. You you can check yourself the effect of excluding or reenforcing the coincidence dip, highering or lowering the mass-spring dip etc. [/quote]

Eric, thank you VERY much - that was very clear and understandable (even by ME :? ) I've got all your Excel files on my laptop and hope to find the time this weekend (if the evil Bad Luck Genie leaves me alone long enough) - I'm looking forward to this, it feels like I'm poised on the brink of a breakthrough here ('bout damn time, too) In spite of only understanding maybe 1/8 of the "why", I've gotten pretty close to the mark on best wall construction - now, hopefully I'll reach a point where I can also explain it. Again, thanks so much for your time and patience... Steve

I am right there with yah steve. It makes total and complete sense to me now (ok, I lied 50%) but I do understand the pressure wall explanation. And from sticking my head in the wall cavity I have built here in my studio and hearing all the Low end that was trapped. (I built my walls, cranked the PA then cut the opening for the window and listened to the diffrence. It starts to come to light for me. I actually have about 20" of space between my control room wall and my Live room wall. But it does make sense to me. The damping effect of the space, the walls and the insul. The test will be for me to explain it in extreme laymens terms. LOL. Happy bulding. Bryan Giles

Howdy' The walls are up and the soffits are done. So far with only the mix room walls finished I have to say the rubber thing seems to be working. Very little of anything is getting through that wall. Is it worth it? For the extra $200. I spent I don't know. As for the hastle of holding 65lb rubber sheets in one hand and trying to use the screw gun with the other, I should have called for some help!!! The speakers are in the soffits and so far they sound pretty good. The bass coming through the wall is minimal. I can't wait to get the second wall finished. Thanks all for the tips. Eric