Oh but now you've called him speker dave! Also, why do you say stiffer and more damped in the same sentence as if to suggest thicker MDF does both these things? I thought the point he was making is that thicker doesn't mean more damped :-s
A material will have a certain amount a inherent damping. Thicker will have more of that damping. As well, thicker is oft done by laminating layers. The impedance change across the glue boundary will have a significant damping effect (ie 6 x 1/8" MDF will outperform 3/4" MDF)
dave
Stiffer, more damped, and less resonant are all good things. Its just that there is a little conflict between them.
Here is what we should all agree on. A better cabinet is one where the resonances are low enough in level or far enough away from frequencies that we are sensitive to, so that we just don't hear them. The output we hear should be from the speaker cone and not the cabinet walls.
You weren't right earlier when you said the resonances would be the same no matter the thickness of the material. For a standard (homogenous) material thicker means heavier and thicker also means stiffer. Heavier would push resonances down and thicker will push resonances up. It turns out that the increase in stiffness outruns the increase in weight so thicker means that the net effect is pushing the resonances up.
That in itself is okay, but the debate on the thread is whether pushing resonances up is a cure. Everyone seems to think that either the resonances will drop in level (walls are, after all, heavier and stiffer) or maybe they can be pushed so high that they no longer concern us.
A segue into mechanical systems. Mechanical systems are modeled with combinations of mass, stiffness, and damping (like electrical systems are Ls, Cs and Rs). As an example, a car on its front suspension can be thought of as a body mass sitting on springs with a certain spring rate (pounds per inch) and, of course, there are shock absorbers in there somwhere. Body mass and spring rate alone determine bounce rate. Hit a bump and the car will bounce at a certain frequency. The shock absorbers determine Q of the system, which is how quickly the car stops bouncing after you hit the bump. Have weak shock absorbers and you will bounce a long time after a bump.
Speaker cabinets are the same. The mass of each wall along with the inherent stiffness will form resonances. Corner joints play into this and determine how much the back couples to the side, how each one effects the next one, etc. So we have a pretty complex system with a number of resonances.
Harwood wrote a paper (and others have come to the same conclusions) that we must not ignore the damping aspect. He started by measuring materials and finding that most of them didn't have enough inherent damping to kill the resonances as much as he wanted to. He found that a significant thickness of bitumin (British for tar) or tar paper would do it, but only if he was careful with cabinet design. The problem was that if he made a cabinet heavier (thicker walls) it actually became harder to damp.
This is back to our car analogy. We can make the body heavier and the springs stiffer but we are still going to react to the bumps. As we make the car heavier it turns out that the shock absorbers become less and less adequate in controlling bounce. They would have to scale up in proportion to body weight or spring stiffness. An SUV needs bigger shocks than an econobox.
Harwood found that as the cabinet wall became thicker the damping became less effective. His best overall result was with a moderately thin cabinet (1/2 inch plywood), and a lot of damping material.
We aren't really talking about putting speakers in a shoebox. Cardboard is too light and soft and would hardly present a barrier to sound. But once we get enough mass in the cabinet walls to knock the broadband level down adequately, we need to really deal with the resonances. Measurements of Barlow and Harwood show that thicker walls move the resonances up some but not out of band and the resonance levels don't drop at all. Damping needs to be the solution and we need to find the right compromise between wall thickness and the amount of damping we can apply.
I hope that explains it.
speker dave
Nice.
I take it the silicon glues the rubber to the MDF?
Does the silicon go hard or remain soft? I ask because I suspect you might get most efficient damping if the silicon goes hard.
The harder the boundary between the rubber and MDF, the more constrained the rubber layer is, and thus more damping.
I think it was Sony that got a patent on sandwiching soft vinyl in the sides of speaker boxes and seem to remember the there was discussion of the advantage of the glue being hard and thin.
I appreciate the clarity. Now what I'm wondering is the application to different speaker types. Is there any difference in the details as applied to sealed, ported, and OB types? Or as applied to sub boxes vs. full-range or multi-way enclosures?
Hello,
It looks like the theory discussion has been exhausted. So my thoughts are; what works?
Take the best of your creations and place a cheap transistor radio inside; what sound comes out? What sound comes through the speaker cone? What sound comes out the tuning port? What sound comes out through the cabinet wall?…..
The entire concept is to keep the sound inside the cabinet from coming out to color the sound on the outside.
Oops a little more thought on theory. Composite walls made from materials of completely different materials and yes remember that boundary layer of goop, mastic or whatever you call the adhesive also improves the sound transmission performance. Each layer adds a band pass filter and dampening that eats vibration. The following links are food for thought.
STC Ratings
http://www.custombuildingproducts.com/docs/TDS106_WonderBoard_3_11.pdf?user=arc&lang=en
http://www.henry.com/fileadmin/pdf/current/tds/HE209_techdata.pdf
DT
All just for fun!
It looks like the theory discussion has been exhausted. So my thoughts are; what works?
Take the best of your creations and place a cheap transistor radio inside; what sound comes out? What sound comes through the speaker cone? What sound comes out the tuning port? What sound comes out through the cabinet wall?…..
The entire concept is to keep the sound inside the cabinet from coming out to color the sound on the outside.
Oops a little more thought on theory. Composite walls made from materials of completely different materials and yes remember that boundary layer of goop, mastic or whatever you call the adhesive also improves the sound transmission performance. Each layer adds a band pass filter and dampening that eats vibration. The following links are food for thought.
STC Ratings
http://www.custombuildingproducts.com/docs/TDS106_WonderBoard_3_11.pdf?user=arc&lang=en
http://www.henry.com/fileadmin/pdf/current/tds/HE209_techdata.pdf
DT
All just for fun!
That would only illustrate issues with a small proportion of the problem. The greatest source of energy to excite box resonances is the mechanical coupling between driver & box and generated by the moving cone. Energy from air coupling inside the box can at most be 33% of the total energy (and likely less, and certainly less in any vented box)
dave
I thought this was for the most part a healthy discussion. All sides seem to have done some homework and make valid points. I have no intent on squelching that, I just wanted to add some practical solutions to the theoretical problems. As for example, this statement: "The greatest source of energy to excite box resonances is the mechanical coupling between driver & box and generated by the moving cone." Unless someone disputes this, what could be done to minimize effects from this necessary coupling?
Having an old friend that is a welder, we used to build cabs from 2mm steel. Two layers of steel with appr. 2 mm. of silicone inbetween. And the inside was silicone-glued with different shaped bricks of concrete-injected MDF.. The frontplate was four layers of steel laminated with the same silicone.
The loudspeaker was a closed cab with four Scan-speak 10" in a double isobaric. When driving the speaker to earsplitting levels and putting a finger on whatever point on the cab's surface there was not the slightest vibration.
Peter
The loudspeaker was a closed cab with four Scan-speak 10" in a double isobaric. When driving the speaker to earsplitting levels and putting a finger on whatever point on the cab's surface there was not the slightest vibration.
Peter
One way is to use 2 drivers in a push-push arrangement with the backs of the drivers direct coupled to each other. I estimate (WAD) that with good coupling 90% of the energy can be actively cancelled. I almost always build woofers this way.
dave
SimontY, I would brace that 3 inch MDF wall enclosure to death, that is where the stiffness would come from 🙂 then I'd glue gooey stuff to insides of those braced to death walls, the gooey stuff would deal with high freq resonances that my braced to death box would exibit. I'd have a rather good enclosure. But then the backwave of the cone will be reflected back outside throught my woofer cone 🙁 so maybe no box would be the way to go? But wait, all my favorite music is mixed using box speakers, if I was to listen to these recordings using Dipoles, wouldn't it be wrong? ok, back to shoe boxes 🙂
Btw - it leads me to some conslusion that a given cabinet-construction's response to the pressure inside is very relative to the SPL! Our object was a tremendous pressure inside a closed cab at high levels.A ported cab. running at lower levels don't need as much attention...
Just a thought
Peter
Just a thought
Peter
Veikko, dipoles don't necessarily sound wrong, just different. It's refreshing to hear a sound free of the box, though what you do miss sometimes is the chesty air-pressurisation on loud dance music, which they can't really do (unless maybe it was an enormous dipole woofer array). I like my 24mm ply dipoles but I am thinking about an infinite baffle attic woofer setup to cover the <40hz range. Sorry, that is a bit off-topic!
I was thinking about an attic, before my wife asked to remove stairs and doors from living room to workshop, so I put subwoofer under the floor.
The floor when was covered by carpet sharply resonated on 11 Hz, now after I put floating hardwood on it on 6 mm cork it does not.
The floor when was covered by carpet sharply resonated on 11 Hz, now after I put floating hardwood on it on 6 mm cork it does not.
well designed Dipoles are closest thing to distortion free loudspeaker, agreed. It's just that music that i listen to isn't mixed and used during recording, using dipoles, that's all. If we had perfect recordings, dipoles would be the only way to go.
those speakers PeterDupont described would be really intersting to listen to too. After u hear what good drivers are capabale of if freed of enclosure resonances (even in typical box design), it is impossible to go back to mdf boxes. Curious thing, good overkill box speaker gets very, very close to sound of dipoles. But it takes heroic box to compete with dipoles in midrange, something like PeterDupont described maybe
I've heard a concrete/mdf thing, ScanS paper/carbon 6 inchers and average mid range ScanS tweeters. Cabinets were few hundred pounds heavy, front baffle was basically MDF and maybe 5 inches of concrete, narrow baffles, just enough to hold SS woofers. They sounded unreal. They sounded like there was no box. we had Totem Model 1 in same room, not a bad commercial design, Totems sounded like there was a blanket thrown on them compared to to the ugly concrete things.
those speakers PeterDupont described would be really intersting to listen to too. After u hear what good drivers are capabale of if freed of enclosure resonances (even in typical box design), it is impossible to go back to mdf boxes. Curious thing, good overkill box speaker gets very, very close to sound of dipoles. But it takes heroic box to compete with dipoles in midrange, something like PeterDupont described maybe
I've heard a concrete/mdf thing, ScanS paper/carbon 6 inchers and average mid range ScanS tweeters. Cabinets were few hundred pounds heavy, front baffle was basically MDF and maybe 5 inches of concrete, narrow baffles, just enough to hold SS woofers. They sounded unreal. They sounded like there was no box. we had Totem Model 1 in same room, not a bad commercial design, Totems sounded like there was a blanket thrown on them compared to to the ugly concrete things.
Last edited:
Hmm, very interesting! Sounds like another vote for concrete. I know someone on this forum who recently tried some granite speakers but didn't like them at all. Also quite interesting... wish I'd heard them with my own ears to see what he didn't like. I wondered if they'd sound "slow" due to all the mass, but now I'm not so sure after reading all this positive stuff about heavy boxes.
Harding silicon or not, The effect is negligible. It's the density shifts that kill resonances. The more the better
- Status
- Not open for further replies.