Sub Bracing Info

[icor1031]

Hope you find this helpful.

Originally posted by trev at http://www.diyaudio.com/forums/subwo...subwoofer.html

I edited this to make it readable, etc. and added emphasis.
Unfortunately this forum removed my indents. So, here's an RTF: http://www.havocgaming.com/subs.rtf

Here's a video of B&W's matrix design. http://www.bowers-wilkins.com/Discov...es/Matrix.html

B&W loudspeakers have done a lot of research in to the effects of cabinet material on speaker response, especially regarding the effects of stored energy and panel modes.


I'm sure you're all familiar with young's modulus, which is a measurement of stiffness and is the first thing to look at when assessing a material for the walls of an enclosure.
This modulus shows us that concrete is an extremely stiff material when compared to wood or board and, at first, seems the obvious choice when designing an ultra-rigid enclosure.


But, you have to step back and analyze what the point is of all this rigidity.
When a driver moves inwards, it pressurizes the cabinet and the cabinet will expand minutely and after, when the diver moves back out, the cabinet contracts again; but not instantly.
The box stores some of the energy from being pushed out, and when it's not being pushed by the driver any more, it releases this energy as sound.
The less rigid the box is, the more energy it stores and, consequently, the louder this sound is and the longer it goes on for.


The overall effect on the sound is that the bass sounds "boxy". It will sound loose and wallowy with no definition.
In contrast, a tight, rigid box will sound more defined.
When the bass stops, the sound stops dead with no ringing or over hang.
In general, a more rigid box will sound better in every aspect when compared to a less rigid one.


The point of B&W's experiments were to ascertain which method of increasing the rigidity gave the best results.

The first being to increase the rigidity of the enclosure walls, i.e: making the box out of concrete.
And the second being to increase the rigidity of the box itself by using reinforcement i.e: building the box out of 18mm mdf and using further panels the same size with holes cut in slightly spaced apart, running up and down the cabinet so as to create a honeycomb effect.

They call this Matrix(tm) bracing and use it in most of their expensive speakers such as the N801 monitor.
To understand why they chose such a complicated bracing system, you have to look at whats going on with a panel when it's being excited by a driver.
A totally unbraced panel will wobble around more or less in sympathy with the driver, acting kind of like a terrible ABR or port.
If it was braced with a batten, supported by a another panel like the opposite wall of the box, it would only really stiffen the panel in the region closest to the end of the batten, where it joins with the wood and the energy would avoid the stiff part of the panel and diffract around it and the overall loss in stored energy would be minimal.
You could brace a box with loads of battens and, although it would be physically very strong, because we are dealing with vibration as opposed to just stress, it still wouldn't be very much stiffer than the unbraced version.
A batten is only supported at each end and so has nothing stopping it moving side to side and the join has a very small surface area.
A panel on the other hand, is supported (preferably glued and screwed) on all four of it's sides and offers much more support, but is still capable of lateral wobble and so we use another panel to brace that, and another to brace that, etc. until we end up with a grid of panels (matrix) inside the box.
The end result is a structure whereby every point is supporting and being supported by every other point.

It is because crystals such as diamond have a molecular structure with the same properties as this (every point supporting every other) that makes them so rock 'ard.
Anyway, B&W did some comparison tests between the concrete box and the matrix box, to see which one sounded better.
The boxes were of the same size with same thickness of material (18mm). Each box was placed face down into a huge box of sand, so as to absorb all the energy from the front and a microphone was placed above the back panel, to record the sound output from the box itself.
The results were incredible. They found to start with that with your average hifi cabinet, up to 20-30 % of the sound you were hearing was being made by the box itself, instead of the driver, but with regards to the comparison test, the matrix box was found to store ten times less energy than the same size concrete box.
In other words, it sounded ten times better. The amount of time and money that goes in to build a matrix cab is much more than the concrete box but the matrix has every other advantage.
Using concrete is a bit like cracking a nut with a sledgehammer. There are much easier and better ways to make wicked subs than to resort to such extreme materials.
Using a complete matrix is a bit over the top but the basic principle of the bracing is the most important thing.
If you're not going for a full matrix and just using some panels to stiffen the box, then avoid symmetry when placing the panels.
The more panels you use, the more you will have to increase the box size to make up for loss of volume, but this is only a small figure.
Many manufacturers use things like 22mm birch ply for their bass cabs, but I use 18mm mdf. It's stiffer than normal ply, but not birch because birch ply is like bloody titanium and blunts all your tools and gives you cramps and aches and weighs a ton and costs a fortune (but other than that, it's great) with a bracing system like described and my cabs pee all over professional kit when it comes to sound.
Another BIG advantage to a properly rigid box is the surprising increase in efficiency.
When the driver isn't wasting all that electricity on vibrating the box, it uses it to make sound and the difference is shocking.

There are many ways you can brace a box, but remember that the point is that the smaller the panel, the more rigid it is and by bracing the center of a panel with the end of another panel, you effectively turn that into two panels, each half the size.
A batten support doesn't do this. Use panels to reduce the surface area of other panels and arrange them in such a way so that they are all bracing each other but try to avoid too much obvious symmetry.
My 18" subs are bloody heavy (not like concrete though), but are so rigid that if you touch the sides when they are on full whack, you can't feel any vibration.
Minimal loss, maximum sound.

Remember: PA companies are marketing a product and sound quality is not their primary concern. Don't be tempted to copy what they do because you think it's the proper way to do it because you are spoiled with PA. Big amps, big drivers=loud music.

With hifi, you're limited to small boxes and have to get the most possible out of them which is why all the really cool advancements in speaker technology, that actually make things louder and sound better, come from the hifi world.
PA just use horns (yuk). If you apply this technology to PA the result is another level of quality when compared to Turbopants or some other macho brand.
Dig into hifi technology and you can build subs that make a mockery of professional kit.
The biggest thing that stands out with a really rigid box (with a quality driver of course) is the content. There is so much going on in the world of bass that we don't normally hear from so-called "subs" (subsonic bass means below hearing, sub is not just another word for bass) that it's a whole new world once you start to play tunes through a proper sub.

Concrete subs would look really cool though...

[bkeane1259]

Informative...thanks. My router bits got dull just looking at those braces

That's definite CNC work there.

[diy speaker guy]

If the message you take away from all that is that more bracing is better than less, then fine, but even so there's definitely a point of diminishing returns. Don't try to read more into his confusing block of text than that, a lot of what he says is exaggerated, misleading or just plain wrong, it would take quite a while to correct all of it.

Bill Fitzmaurize regularly posts about bracing, showing a picture of a "window" brace and a picture of dowels. The dowels are just as effective as the window brace (there might actually be a small difference, but it is extremely small). If your side to side, top to bottom and front to back dowels connect in the middle, and you use enough dowels to adequately support the surface area of the panels, that's all you really need.

I didn't see any pictures of this super matrix bracing (can't watch videos on dial up) but it sounds like a big waste of time, money and precious enclosure space, although I'm sure it looks cool.

Also, he makes no mention of density vs frequency, which ideally should be considered when making the selection of material to use in the first place, but it's clear he doesn't know much about the difference between ply and mdf anyway.

[icor1031]

If the message you take away from all that is that more bracing is better than less, then fine, but even so there's definitely a point of diminishing returns

I think he hints at this, here:

Using a complete matrix is a bit over the top but the basic principle of the bracing is the most important thing.
If you're not going for a full matrix and just using some panels to stiffen the box, then avoid symmetry when placing the panels.

What I took from it is basically, try to use asymmetrical braces that support each other, and the center is more important than corners etc (the last I already knew.)

[diy speaker guy]

What I took from it is basically, try to use asymmetrical braces that support each other, and the center is more important than corners etc (the last I already knew.)

Good work, I think you hit all the important points in a single sentence.

[bkeane1259]

I'll save ya the trouble:



I think it's pretty interesting, but I'll never do it. I'll bet the cabs are dead as all heck.

On the flip side.....

If you want some good entertainment, next time you're at Best Buy, take 5 minutes to do the knuckle rap test on the various stuff they have on display at Magnolia. You'll be amazed at most of the results.

[icor1031]

So, here's a question...

What happens when you use 2 PRs? You have 4 of the 6 walls being used: 2 by a PR, 1 by the driver, 1 by the plate amp..

Do the devices act in the place of bracing, or is just going to sound relatively poor?

[diy speaker guy]

So, here's a question...

What happens when you use 2 PRs? You have 4 of the 6 walls being used: 2 by a PR, 1 by the driver, 1 by the plate amp..

Do the devices act in the place of bracing, or is just going to sound relatively poor?

Assuming the box is very small, all you would need is top to bottom bracing. All the other 4 panels are mostly filled with cone area or amp, and as you already said yourself, the middle of the panels are the most important to brace. These panels don't have a middle, the box you described is mostly just corners and barely enough baffle to hold the components. Brace what you can and don't worry about the rest.

[icor1031]

Ironic, these guys might be able to save money by using PRs instead of that matrix design.

Any way, thanks. That's what I want to do, add two PRs to get a lower response.

Assuming the box is very small, all you would need is top to bottom bracing. All the other 4 panels are mostly filled with cone area or amp, and as you already said yourself, the middle of the panels are the most important to brace. These panels don't have a middle, the box you described is mostly just corners and barely enough baffle to hold the components. Brace what you can and don't worry about the rest.

[diy speaker guy]

On the flip side.....

If you want some good entertainment, next time you're at Best Buy, take 5 minutes to do the knuckle rap test on the various stuff they have on display at Magnolia. You'll be amazed at most of the results.

Thanks for the pic.

The knuckle rap test is very misleading. Since a knuckle rap is not bass it doesn't tell you much about enclosure performance at bass frequencies. Massive, dense and heavy might sound good in principle and it might make a knuckle rap test sound pretty dead, but massive, dense and heavy is what it takes to store energy at bass frequencies. If you make the box less massive, less dense and less heavy and very well braced, it will sound worse in the knuckle rap test but it will actually store less energy at low frequencies than if it were big and heavy. For a mid/tweeter enclosure (that doesn't have to play bass) massive, dense and heavy is perfect, since the speaker won't be playing the bass frequencies that will excite the panels at all. If you match up the right materials (and thickness of materials) to the frequencies the speaker has to play, there are opportunities to eliminate energy storage as much as is possible.

[billfitzmaurice]

I'll save ya the trouble:



I think it's pretty interesting, but I'll never do it. I'll bet the cabs are dead as all heck.

It works well I'm sure, but it's massive overkill. I'd say fully 3/4 of the material employed is unnecessary, that the same level of functionality could be realized with 1/4 the material used. That would make the cabs far lighter, and less expensive. Not the expense matters to BW. IMO the main benefit to that scheme lies in its marketing value, and since BW heavily touts it they obviously agree.

[duanebro]

So how much bracing is needed? And at what point are the walls thick enough? Are curved walls more dead? I am trying to come up with a cabinet rebuild because I didn't make the first one dead enough. So the Big question:

At what point does it become overkill?

[mattp]

So how much bracing is needed? And at what point are the walls thick enough? Are curved walls more dead? I am trying to come up with a cabinet rebuild because I didn't make the first one dead enough. So the Big question:

At what point does it become overkill?

In my opinion it really doesn't become overkill. Curved cabinet walls are generally stronger and help reduce internal standing waves but they are time consuming to build. I personally like to brace the heck out of my cabinets (think Lunchmoney or Joeybutts style) but I usually use scrap wood as side to side braces. As long as you consider the volume reduction which usually isnt much you will be fine.

Here is a screen shot of my next buld with the bracing shown. Overkill for some, but if it wasn't there I wouldn't be happy. Driver and bracing displacement is .374 cubic feet.

[lunchmoney]

Very surprising that there's no mention of what range of frequencies will be audible from panel resonance.

A while back I made a very heavily braced sub (which sounds incredible tight and clean), only to have some knowledgable folk say:

"hey nice job, but bracing a sub like that is a waste of time because you won't hear resonance at those frequencies"

Ever since I've been scratching my head at whether I pointlessly created a ridiculously heavy box or not.

[mattp]

Ever since I've been scratching my head at whether I pointlessly created a ridiculously heavy box or not.

IMO you didn't. As far as which frequencies will be audible, this is interesting and something I noticed about my Apaches build. I braced the h*ll out of the woofer cabinet and still had some resonance problems, at least by feeling the box. Whether it was audible or not I'm not sure. The upper mid cabinet was a different story. It vibrated within a pretty wide bandwidth and it did seem to have a mellow ring to it. I only used one small brace thinking a mid enclosure wouldn't need much but I was wrong. I know feeling the cabinet vibrate only says so much but it has to amount to something in the long run, negative or not. Me no likey.

[billfitzmaurice]

So how much bracing is needed? And at what point are the walls thick enough? Are curved walls more dead? I am trying to come up with a cabinet rebuild because I didn't make the first one dead enough. So the Big question:

At what point does it become overkill?

I've found that with curved walls made of 1/4" plywood spans of six to eight inches between brace points are adequate. I have electric bass cabs with 1/8" BB curved walls braced at 4 inch intervals that are solid as a rock.

[duanebro]

Any DIY software for modeling this sort of thing?

[duanebro]

I've found that with curved walls made of 1/4" plywood spans of six to eight inches between brace points are adequate. I have electric bass cabs with 1/8" BB curved walls braced at 4 inch intervals that are solid as a rock.

How many layers of 1/4" did you use? Do you have any pics?

Thank you,
Duane

[lunchmoney]

IMO you didn't. As far as which frequencies will be audible, this is interesting and something I noticed about my Apaches build. I braced the h*ll out of the woofer cabinet and still had some resonance problems, at least by feeling the box. Whether it was audible or not I'm not sure. The upper mid cabinet was a different story. It vibrated within a pretty wide bandwidth and it did seem to have a mellow ring to it. I only used one small brace thinking a mid enclosure wouldn't need much but I was wrong. I know feeling the cabinet vibrate only says so much but it has to amount to something in the long run, negative or not. Me no likey.

Well it certainly didn't hurt, that's for sure. It sounds incredibly tight and clean. It just seemed like a good idea, particularly since the sub is sealed. And of course knowing the psychology of these things, had I NOT braced it so much, I always would have wondered if I should have

[diy speaker guy]

Very surprising that there's no mention of what range of frequencies will be audible from panel resonance.

???

I've mentioned it already a couple of times. Or are you talking about the original post?

"hey nice job, but bracing a sub like that is a waste of time because you won't hear resonance at those frequencies"

They probably meant that any audible frequencies would blend into the mix down at sub frequencies, otherwise this statement just doesn't make much sense. If anything, the added mass of all the bracing will acctually INCREASE energy storage (resonance potential) at sub frequencies.

Remember, it's panel resonance we are trying to avoid if possible. Heavy, dense structures resonate at lower frequencies than light, strong structures.