Home Theater Forum and Systems banner
1 - 20 of 64 Posts

·
Registered
Joined
·
62 Posts
Discussion Starter · #1 ·
At 100 Hz, as long as the biggest box dimension is < 1.7 m (5.6’), potential panel resonances will be above the panel sizes.

In that situation, is there a benefit in bracing?

Thanks
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #3 ·
per SL, the first internal air cavity resonance occurs when the internal length equals 1/2 wavelength. http://www.linkwitzlab.com/thor-design.htm
100 Hz has a wavelength of 3.43 m, half a 100 Hz wavelength is 1.7 m.

I’ve seen numerous posts and web pages on stiffer boxes, two layer construction, heavy bracing, etc. But if the above is true, in the frequency ranges of subs, what benefit do stiffer boxes etc give?
 

·
Registered
Joined
·
866 Posts
I see what you are saying. Everything has natural resonances. What you quoted is for the resonance of the space inside the sub. The panels themselves also have a natural resonance. What you don't want is the panel itself to be resonating with the notes a speaker will play.
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #5 ·
Thanks for clarifying, Boomie

Wonder why Linkwitz doesn’t worry about panel resonances at all, for a sub designed to compliment his “aim for the best”, widely admired etc Orion . . ?
 

·
Registered
Joined
·
866 Posts
Thanks for clarifying, Boomie

Wonder why Linkwitz doesn’t worry about panel resonances at all, for a sub designed to compliment his “aim for the best”, widely admired etc Orion . . ?
That box is about 12" x 12" x 20" and has edge bracing. That isn't super-big. It might just not need anything else.
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #7 ·
Makes sense, thanks.
I searched here on bracing, and didn’t find a huge amount.

One point was, re a 3/4" MDF box: the “general rule of thumb” using MDF for a sub box is brace every 16 inches or so.

Plywood is stiffer than MDF, but would every 16 inches or so be a good rule of thumb for 3/4" ply?
 

·
Registered
Joined
·
1,001 Posts
All boxes using a single layer of plywood or MDF don't need the same amount of bracing. It depends on the driver used. Some drivers don't have very big magnets and don't produce a lot of ft/lbs of energy. Smaller drivers, 6" , in larger boxes is a good example. Drivers with small vas which don't require a large box will perform well without too much bracing.
But large drivers with large magnets and huge xmax in a "large low tuned box" need a LOT of bracing. 16" squares for an 18" driver in a large box, 6 cubic feet or more, leaves too much material unbraced for THX performance. That kind of box will contribute spurious sounds while under large SPL stress. Whether or not these spurious sounds become audible and noticeable to the listener is another question. You sure can feel the box deflections with your hands and fingers! Sonotube enclosures are good for minimizing enclosure flexing but even they are not totally inert. Ported enclosures don't flex as much as sealed.
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #9 ·
Makes sense . .
I have a few bass/ subs on my “to do” (for 3-4 different rooms & situations)
Too many ;) , but I’ve accumulated the drivers, recently got all the plywood cut (all ¾” marine ply); just this week I worked out rebating . .

I have been focussed on driver quality and alignments, and hadn’t thought much about bracing, until recently, especially a couple of nights ago when I read (posted from Navigating Speaker Design: Listening to Walls by Mark Wheeler, Speaker Builder 8/99): “Bracing effects were similar in orders of magnitude to driver selection” - which astounded me – glad I read it before putting the boxes together!!

After all I’ve now read, Dave Dlugos’ braces shown in the drawings for the CSS 15” http://www.creativesound.ca/pdf/sealed-SDX15-131107.pdf (which I found by searching on bracing here), to me seem to be what’s about best, as they also spread the Newtonian energy of the driver as evenly throughout the box as possible, and a 15” at say 105 db and 20 Hz and less must have a lot of Newtonian energy . .
Though like you I don’t like the idea that parts might eventually assume the tensioned shape and become permanently curved. Having to regularly replace the baffle has me wondering . . Dave if you see this and could clarify which panels might slowly warp – the internal vertical brace??

It WILL take a while getting to building ALL of them, but here they are, followed by my guesses of appropriate bracing based on what I’ve read/ just learning/ you’ve said:

Peerless XLS 12” 40 litre sealed sub, with a transform or EQ, my 90% size Linkwitz Thor. Brace: per SL’s Thor - along the long ‘edges’?

Peerless 10” 40 litre sealed ‘midbass’, with a Linkwitz transform or bass EQ around 30 Hz. Brace the same as above.

Peerless XLS 12” in an 80 litre EBS (for HT, it models reasonably well). Dims 12 * 17 * 35” (310 * 430 * 900). (I seriously considered the EBS in Sonosubs, but space/ SAF worked out better for a “coffee table EBS”).
As the EBS is “skinny and long” brace: lengthwise once, and across say twice (per the holey brace).

High efficiency PHL 15” vented midbass (100 – 500 Hz); and the Rythmik 15”, sealed with the servo amp.
Both in identical net 80 litre boxes, 17 * 17 * 23 (430 * 440 * 570).
As the boxes are near to the cube proportions of the CSS 15 (also with a volume close to the middle suggested option for that driver), again I might clone (I should say do similar) the "CSS bracing" too.

Opinions welcome (apart from the obvous that I have to much to do!) :) . .
 

·
Registered
Joined
·
999 Posts
Never hardly can brace too much :bigsmile:

at the start


and two internal views with the window braces



I also added many short brace pieces from front (driver side to back terminal side) among the window braces.
The idea is to make the enclosure as stiff as possible -- thereby raising any resonances to high order freqs.

:whew:
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #11 ·
Thanks Bob, nice work. :)

How did you get a good tight fit of braces to the box sides: cut the braces say 1/32" over size, then somehow (a flush trim router?) trim them down?


To get a sense of relativity:
Two identical size 3 cubic feet boxes, both with 15” drivers:
1. Sealed, 15 Hz – about 80 Hz,
2. Vented, about 80 – 400 Hz (ie not a sub)
Before bracing, the internal standing waves correspond to frequencies of 320 – 420 Hz.

. . while more is better, there are only so more hours available: which should need more bracing?

Thanks again
 

·
Registered
Joined
·
999 Posts
How did you get a good tight fit of braces to the box sides: cut the braces say 1/32" over size, then somehow (a flush trim router?) trim them down?
Measure twice, before saw guide cut, and then still have it a bit off. Used a lot of yellow wood glue -- applied multiple times to all the ends/cracks of the internal braces. :whistling: :blush:
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #14 ·
Putting aside the internal standing waves, and thinking about the Newtonian forces:
For the same dB, when Hz is halved, required cone movement rises 4 times.
Cone movement = Newtonian force.

So a sub that we want flat to 20 Hz, compared to a midbass box with a highpass of 80 Hz, will have 16 times the force to cope with.

So if I conclude correctly, the midbass probably only meeds moderate bracing, maybe say 9? mm (3/8 inch) thick?,
while the sub by comparison . . for higher dB, metal rods wouldn’t go astray . .

I believe some have used iron, steel, or stainless steel pipe, or large threaded rods . .

Any comments or experience?

Thanks
 

·
Registered
Joined
·
1,001 Posts
a midbass box with a highpass of 80 Hz

So if I conclude correctly, the midbass probably only meeds moderate bracing, maybe say 9? mm (3/8 inch) thick?,
while the sub by comparison . . for higher dB, metal rods wouldn’t go astray .
I believe some have used iron, steel, or stainless steel pipe, or large threaded rods. Any comments or experience?
Thanks
How big can a midbass box be using 9 mm material? I think a reasonable size limit would be in the 1-2 cubic feet range with bracing. I've used wooden dowels or (stiff wood) IE: oak, hickory, etc. for braces. Wood so easily glues to wood compared to using metals with wood. But if you've got metal on-hand, it can be a substitute.
I have a pair of JBL CM42 9"x6"x5" made out of metal and plastic. they are very substantial enclosures.
 

·
Banned
Joined
·
1,131 Posts
Putting aside the internal standing waves, and thinking about the Newtonian forces:
For the same dB, when Hz is halved, required cone movement rises 4 times.
Cone movement = Newtonian force.

So a sub that we want flat to 20 Hz, compared to a midbass box with a highpass of 80 Hz, will have 16 times the force to cope with.

So if I conclude correctly, the midbass probably only meeds moderate bracing, maybe say 9? mm (3/8 inch) thick?,
while the sub by comparison . . for higher dB, metal rods wouldn’t go astray . .

I believe some have used iron, steel, or stainless steel pipe, or large threaded rods . .

Any comments or experience?

Thanks
It seems you are over thinking this quite a bit.

When you brace a speaker or subwoofer the main purpose of it is to push any resonant modes out of the passband more doesn't matter, but less can possibly cause audible resonance. Due to a variety of factors subwoofers need minimal bracing compared to something like a midrange to have the resonance modes pushed beyond the realm of interference.

Take my build for example. If you scroll through the images you will see an oak matrix used to brace the subwoofer. This along side the bracing used in the side walls was more than sufficient to make the cabinet audibly inert for the frequencies being played (crossover 70-80Hz). In fact, you will see the two pieces of the matrix are spaced unevenly. Once attached to the cabinet it seems the larger gap causes a resonant mode around 110-125Hz which with the crossover system being used is well out of range from being detrimental. The smaller gaps first resonant node is likely around 200-250Hz due to the closer spacing of the bracing.

No steel or heavy bracing is needed in a typical subwoofer in fact it would be simple and non-detrimental to use less bracing than I did in my specific design while still having no resonance in the passband. If you are really worried about resonance I would suggest you read The Modification of Timbre by Resonances: Perception and Measurement JAES Volume 36 Issue 3 pp. 122-142; March 1988.*

If you get around to building speakers thats when you really would want to worry about maximizing bracing efficiency if the goal is to completely remove all audible resonance from the system. Very few commercial loudspeakers are able to boast this due to manufacturing costs, weight etc...This is a link to a thread where a friend of mine discusses a pair of his computer speakers which are over braced by most standards, but still have audible resonance at certain frequencies.

*There has been other work done on resonance as well, but it seems conclusive that coloration due to resonance is perceived as a negative when rating loudspeakers as shown by credible perceptual research.
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #17 ·
Phil,

> How big can a midbass box be using 9 mm material?

The midbass wood is already cut from *18 mm ply to form a box net 80 litres, 2.8 cubic feet (it’s a 15” midbass).

The 9 mm was my suggestion for bracing, not yet cut, of moderate strength compared to a sub (given that it won’t be playing below 80 Hz, so cone movement is around 1/16th that of a sub).

> Wood so easily glues to wood compared to using metals with wood.

Yes, though I wasn’t thinking of *gluing the metal rods.
I’m just starting to think of exactly how to do it, so something like threaded rods - in one side of the box, out the other.

I think some of the Lambda users do this. Nuts to tighten them up, washers to (allow tightness without damaging the wood) and seal the openings.

They’d be lighter and probably stiffer than wood, possibly cheaper and quicker to build & install.
On the downside, they probably couldn’t lock together, nor spread the driver’s energy from the back of the magnet, like Dave’s do. And you’d probably want to find some sort of caps to cover the exposed outside nuts, unless the Frankenstein look appeals ;)


Maybe a good compromise would be a vertical side-to-side brace of wood, up against the back of the magnet to spread the driver’s energy; and threaded rods for the rest of the bracing.

If you wanted vertical rods, to give clearance under the bottom nuts, you’d need ‘feet’ under the box, but rubber feet, say 15 mm high could be a good idea (prevent walking) anyhow.
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #18 ·
Andrew,

> work done on resonance as well, but it seems conclusive that coloration due to resonance is perceived as a negative when rating loudspeakers as shown by credible perceptual research

Yes, so as I’ve recently begun to appreciate fully. . I looked at your Infinity project – whoa! VERY impressive.
I don’t however have the time or skill to aim for something that substantial, at least for the foreseeable future . .

> the main purpose of it is to push any resonant modes out of the passband more doesn't matter, but less can possibly cause audible resonance

More only matters that’s its “unnecessary” materials and time, so I’m just trying to learn what’s appropriate and efficient for different situations (sub/ midbass/ mids).

I don’t expect to remove all resonances, say at least 80 or 90%, but not spend time “over-building” in the range of diminishing returns. I don’t have too much time for building the perfect box, mostly just when I take leave, and I have a number of projects I want to do.

So personally I’m aiming for bracing that’s appropriate for the box’s use, not eg twice the bracing for a further say 10-15% less box resonances. That’ll have to wait until I retire . .

> Due to a variety of factors, subwoofers need minimal bracing compared to something like a midrange to have the resonance modes pushed beyond the realm of interference

So the higher cone movement *isn’t as big an issue as the resonances being e.g. in the midrange’s passband . .

> If you are really worried about resonance I would suggest you read The Modification of Timbre by Resonances: Perception and Measurement JAES Volume 36 Issue 3 pp. 122-142; March 1988.*

Thanks, I will try to get that at the nearest big technical library.

> No steel or heavy bracing is needed in a typical subwoofer

Unless I’m missing something, metal rod bracing vs wood seems a good shortcut & trade-off; but (my box designs inc calcing the internal standing waves) looks like it’s really only worthwhile in a midrange . .
 

·
Banned
Joined
·
1,131 Posts
> Due to a variety of factors, subwoofers need minimal bracing compared to something like a midrange to have the resonance modes pushed beyond the realm of interference

So the higher cone movement *isn’t as big an issue as the resonances being e.g. in the midrange’s passband . .
The midrange is a completely different beast when compared to a subwoofer. Exponentially more thought and bracing would go into making a fully inert midrange cabinet as compared to a subwoofer. For most subwoofers if one connects the XYZ axes and places bracing horizontally every 15" or so the cabinet will have resonance bands far beyond passband. For a completely inert midrange far more dense (2" spacing on the XYZ) bracing must be used while implementing a constrained layer of dampening material such as Peal N Seal. *Please note these are just examples and should not be thrown in random builds as I have not outlined the method exactly.*

This is a good example of another Kappa Perfect sub build that will be inert in the passband as well that is far less complicated than mine.

> No steel or heavy bracing is needed in a typical subwoofer

Unless I’m missing something, metal rod bracing vs wood seems a good shortcut & trade-off; but (my box designs inc calcing the internal standing waves) looks like it’s really only worthwhile in a midrange . .
If you properly tie together the axes you will make the use of metal in this application complete and utter overkill. Now if you were making a midrange and able to weld I would be encouraging you to build a steel matrix as this would be an extremely efficient way to brace a midrange (not in terms of weight). In fact I wish I had welding experience as that would be exactly what I would do. You really do not need to use more than Oak for your bracing, just think it through and post pictures of your plans so comments can be made.
 

·
Registered
Joined
·
62 Posts
Discussion Starter · #20 · (Edited)
> The midrange is a completely different beast when compared to a subwoofer. Exponentially more thought and bracing would go into making a fully inert midrange cabinet as compared to a subwoofer.

Can you point me to a source with more info on that?

> For most subwoofers if one connects the XYZ axes and places bracing horizontally every 15" or so the cabinet will have resonance bands far beyond passband.
For a completely inert midrange far more dense (2" spacing on the XYZ) bracing must be used while implementing a constrained layer of dampening material such as Peal N Seal.
*Please note these are just examples and should not be thrown in random builds as I have not outlined the method exactly.*

15" spacing for subs is a similar starting point to another post I read.
Linkwitz recommended 4" spacing, but I don't know in what context . .
Having said that, what is the method you refer to?

> If you properly tie together the axes you will make the use of metal in this application complete and utter overkill.

I don't think I'll ever be welding capable - for midranges, I'll ask a friend who's a metal worker if he knows of an off the shelf fitting that allows 2 (threaded rods) at 90 degrees to tie into each other.

> a steel matrix as this would be an extremely efficient way to brace a midrange (not in terms of weight)

I'll check what suitable moderately priced materials are stronger for their weight. Why do you think a metal matrix would weigh more than a wood matrix of the same strength?

> You really do not need to use more than oak for your bracing

Here we don't have oak, though several other hardwoods.
However as outlined above, it seems to me that metal rods have net advantages, especially in being quicker to build & install.

That could allow more bracing in the same amount of time ;)

> Post pictures of your plans

There are no CAD/plans (just Unibox results in a spreadsheet of box dimension scenarios). I'm attempting to weigh up and work out how to implement a basic appoach that I saw mentioned on the Lambda forum, with no details.

A drawing could show desirable spacing for the passband, which is becoming clear.

If the basic approach isn't clear, I'll redraw my sketch more legibly, scan & post, if that helps convey it.

Thanks
 
1 - 20 of 64 Posts
Top