Bracing again - Page 5 - AVS Forum | Home Theater Discussions And Reviews
Forum Jump: 
 87Likes
Reply
 
Thread Tools
post #121 of 337 Old 02-11-2016, 01:26 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,560
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 587 Post(s)
Liked: 540
Quote:
Originally Posted by Bill Fitzmaurice View Post
+1. The most interesting aspect of the USS Zumwalt isn't its use of stealth technology, though. Scuttlebutt has it that it's a test platform for next gen weaponry. It's capacity for generating electricity seems to be quite a lot higher than what's required for propulsion and all of the other conventional shipboard power needs. It will probably be the first ship of the line to be equipped with lasers and rail guns.

Bigus is offline  
Sponsored Links
Advertisement
 
post #122 of 337 Old 02-11-2016, 01:50 PM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5768
Quote:
Originally Posted by Bigus View Post
And the Captain's name is James Kirk.

Arm photon torpedoes.
Bill Fitzmaurice is offline  
post #123 of 337 Old 02-11-2016, 02:27 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,560
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 587 Post(s)
Liked: 540
OK, I got bored for a minute and did the simple case comparison of a rectangle cross section used in its stiffest direction as a panel beam stiffener vs stanchion, ends fully constrained (true only if 90 deg connection to next panel and stiffener is perfectly rigid), mid span deflection vs stanchion to opposing panel of same length and cross section as beam stiffener.

Using beam MOI (bh^3)/12, and setting deflection equivalent and finding needed relative weights to reach that deflection (W1, W2):

[(12)(W1)L^3]/(E)(192)bh^3 = (W2)(L)/bhE

We get W2/W1 = (1/16)L^2/h^2

The results depend on the cross section and length we're talking about. Actually, they really only depend on the length and height of the beam (and it being equivalent to one dimension of the stanchion), the width of the base of the beam cancels out. So a 2x4 vs a 1x4 has the same result, which I find interesting. Makes sense when I think about it, but wasn't obvious at a glance.

Taking a 2x4" stiffener, 24" in length, deflection is 2.25 times greater used as a beam along the panel compared to a stanchion to opposing panel.

There are some dimension which will push the beam towards less deflection than the stanchion. Specifically, if the span becomes short enough relative to the size of the stiffener. For example, using the 2x4" cross section and making the box a 1 foot cube internally, the beam actually deflects only a little over half that of the stanchion. But that's a fairly impractical scenario. Or, if the height of the stiffener becomes large relative to the beam. Like using a 1x10" stiffener in a normal sized box.

Consider something more realistic, like a 2 foot per side cube box, and our stiffener/stanchion is more like 1x1". The deflection is now 36 times greater when used as a beam vs a stanchion!

But this does tell us one thing... if you're willing to make the stiffener really deep (like maybe a well designed window/shelf brace where cutouts don't undermine the beam depth, stiffening can still be quite good.

Easier to just use stanchions.
ronny31 likes this.


Last edited by Bigus; 02-11-2016 at 02:30 PM.
Bigus is offline  
Sponsored Links
Advertisement
 
post #124 of 337 Old 02-11-2016, 02:54 PM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
Sweet job! But I've already familiarized myself with Autodesk Fusion 360 and done two dozen simulations with various designs and materials using the pressure made by the law of perfect gases if the 600x600x600mm enclosure volume was to be reduced by the amount of air moved by a beefy 18" subwoofer cone moving inwards with 20mm xmax in a closed enclosure. Also did some modal frequency simulations and checked some ballpark figures on what db is made by the area of material moving at those frequencies with such and such displacement due to pressure stress. Full-range speakers got a lot more complicated to make well, and a lot less complicated to make well, at the same time.
Autodesk Fusion 360 is free for 1 year btw. Just register as a hobbyist.
ronny31 is offline  
post #125 of 337 Old 02-11-2016, 07:51 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,560
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 587 Post(s)
Liked: 540
I've used autodesk products for about two decades, but to be honest I've never played with fusion 360. I did a career uturn about a decade ago, and use their drafting and solid modeling tools now when needed, which is really just occasionally for more difficult speaker designs.

So I'm not sure what the capabilities of 360 are. But I'll give advice in general terms in case these things aren't obvious. Stress analysis isn't necessary, forget about that part. Modal analysis is where its at for speaker enclosures. Resonance of the panels. About the best you can hope for is moving the resonance out of the passband. And if you can do that while adding mass, you know you're reducing amplitude at resonance. Beyond that, I have my doubts about autodesk software being able to accurately solve for amplitude given a specified input force driver, even a simple one an impulse. Let alone a sine wave of a single frequency. Even some of the most powerful commercial software struggles with that, mostly because the viscoelastic damping effects are often nonlinear and notoriously hard to model adequately. Our R&D group used software we designed in house and extensively vetted against experiment to solve a single class of problems... machinery isolation rafting on submarines tonprevent noise transmission... and that was some beast of a software. Expecting to get reliable results from a general purpose solver seems optimistic... but then again I have been out of the game for a while.

If you're going that far, just move resonance out of passband while adding mass and especially stiffness, damp if you can, and you're good.

Or, just follow Bill's guide to bracing. Incredibly simple and works as good as is needed, and then some.
Bigus is offline  
post #126 of 337 Old 02-11-2016, 08:12 PM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
Quote:
Originally Posted by Bigus View Post
Modal analysis is where its at for speaker enclosures. Resonance of the panels. About the best you can hope for is moving the resonance out of the passband. And if you can do that while adding mass, you know you're reducing amplitude at resonance.
When you say adding mass, do you mean adding mass to the cabinet? Why does mass reduce the amplitude at resonance or how should I think about this? Thanks
bob_m10 is offline  
post #127 of 337 Old 02-11-2016, 09:02 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,560
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 587 Post(s)
Liked: 540
Quote:
Originally Posted by bob_m10 View Post
When you say adding mass, do you mean adding mass to the cabinet? Why does mass reduce the amplitude at resonance or how should I think about this? Thanks
Adding mass to the panel.

Stiffening the panel drives resonance frequency up, and decreases amplitude of forced vibration even at resonance. Driving resonance above the passband is a pretty good idea for sub enclosures (often not practical for fullrange cabinets), as it moves the most problematic frequency to a range where there is little energy left to excite it. You're then left with dealing with non-resonant vibrations.

Adding mass to a panel (wall of a box) drives resonance frequency down, so for subs, you have to balance that with the amount of stiffness you have achieved. But adding mass also increases the amount of energy needed for an impulse to produce a vibration of a certain amplitude. Since real music is typically varying rapidly in frequency, and you have hopefully moved resonance above the passband, adding mass reduces the amplitude of panel vibration at other frequencies.

The other variable is damping. This removes energy from the panel as it vibrates, and is particularly effective at resonance. This is why damping is quite helpful for fullrange enclosures, less so for subs (but does still have some effect even out of resonance, and can help to quickly quiet any panel resonance that is excited by harmonic distortion products or lower level fundamentals not far above xo point if resonance isn't that high, so it still plays an important role).

Hopefully that makes sense. For subs, the general idea is simple. You want stiff panels that are as thick (massive) as is reasonable given practical budget, weight, and final enclosure size concerns and that doesn't lower resonance into the passband.

Calculating what that is isn't particilarly easy. Using common thickness MDF or quality plywood with BFM's suggested bracing will get you there in most any reasonable enclosure size, so that is easy.
Bigus is offline  
post #128 of 337 Old 02-11-2016, 09:55 PM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
Quote:
Originally Posted by Bigus View Post
Hopefully that makes sense. For subs, the general idea is simple. You want stiff panels that are as thick (massive) as is reasonable given practical budget, weight, and final enclosure size concerns and that doesn't lower resonance into the passband.
Lets see if I understand, can I think of a panel on the side of my sub as guitar string? If we add more mass to the string the note will play lower as it resonates. If we tighten the string (stiffen) the note will be higher. if we hold down half way up the string, (brace) the note will be higher as well?
bob_m10 is offline  
post #129 of 337 Old 02-12-2016, 01:23 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
Adding mass to a guitar string is different to a wood material, most materials get stiffer if you add mass.

About Fusion 360, its a tool for modelling and stress-testing for pretty much anyone who could want to get the product made with CNC machines or 3D printers. As such they are giving it away for free for hobbyists and startup companies so that when they earn some money from it they go with the programs that are familiar to them already, being autodesk software. So compared to six years ago, let alone ten, its quite beastly in capabilities.
To simulate the force a driver puts on the internal surfaces simply calculate the change in gas pressure between neutral position of the cone, and complete inward position of the cone, using the law of perfect gases and the air moved by the cone at xmax. Then apply that force in MPa to the surfaces you simulate (as opposed to a single point force). The trick is to design the test-bed properly, because the anchor point surfaces you use will artificially strengthen the enclosure (those surfaces are held in place by the Hulk).

I'm more into full-range speakers than subwoofers. Subwoofers can quite easily be braced enough, relative to full range speakers. Preliminary results show that full range speakers can easily require a lot more strength than I previously realized. The trick is to not have the reinforcement take up too much volume, I want a "normal" size to my 15" woofer speakers with 12" horn so every liter of bracing is going increase its size more than I really want.
ronny31 is offline  
post #130 of 337 Old 02-12-2016, 07:26 AM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,560
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 587 Post(s)
Liked: 540
Quote:
Originally Posted by bob_m10 View Post
Lets see if I understand, can I think of a panel on the side of my sub as guitar string? If we add more mass to the string the note will play lower as it resonates. If we tighten the string (stiffen) the note will be higher. if we hold down half way up the string, (brace) the note will be higher as well?
Yes, that's exactly right. The panel vibrates very much like a guitar string. Bracing a panel is just like putting a finger on a fret, shortening the vibrating length of the string.

The only thing to keep in mind is that there are many solutions that give the same resonant frequency, but not all behave the same under a given input force. The high string on the guitar (I play brass instruments, not string other than the piano!) might resonate at 330Hz (Google is my friend), but so might a 1/4" diameter steel rod of a certain length. Both will sound the same note when played, but given the same "pluck" from your finger, which do you think will vibrate at a greater amplitude (and thus louder)? Going further, it may be that a certain huge steel beam in a skyscraper also resonates at 330Hz. Do you think it will move much with a typical finger pick??

So while multiple beams or panels might vibrate at a given frequency, we want the one, within reason, that will vibrate at the lowest amplitude at that frequency for the forces transferred to it by the subwoofer or other speaker driver. You can increase resonant frequency by drastically dropping weight, but you get something "flimsy" like a guitar string. You can increase resonant frequency by increasing intrinsic panel stiffness and bracing it to reduce the span (string length). We want to do that.

Distributed or point mass along a string or panel will reduce resonant frequency, but that's mostly OK because the extra mass means there is less reaction to a given input force. And, so long as the mass is used in the panel or bracing itself, it is significantly contributing to stiffness which will be the dominant factor. So long as you don't add "loose" mass that drops the resonance into the passband of the sub, it's all good.

There are even some applications though where that is used, like mass loaded vinyl for sound absorption. In that case we assume we aren't going to move resonance out of the intended frequency range of use, so we just want to make the panel react as little as possible to the energy put into it. Adding mass helps accomplish that.
bob_m10 and ejh2854 like this.

Bigus is offline  
post #131 of 337 Old 02-12-2016, 08:59 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
One hazard with CAD software, you're sitting there minding your own business just trying to make a simple test-sample consisting of two 1220x2440mm 9mm thick sheets of plywood, with 62 and 124 strips between in alternating direction. Before you know it you end up with 7874 contacting faces and your normal (albeit 8 core) CPU can't handle simulating that unless you've got all night... So... Yeah... Might have to merge those two layers together, so that they aren't 62 and 124 separate strips of plywood contacting each other 7688 times. Now to figure out how to merge 186 components.
ronny31 is offline  
post #132 of 337 Old 02-12-2016, 09:23 AM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,560
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 587 Post(s)
Liked: 540
Solid (volumetric) elements are the wrong approach for this sort of problem. A combination of plate elements and beam elements are perfectly adequate and will actually allow for a greater degree of resolution given a set amount of CPU resources and time.

Bigus is offline  
post #133 of 337 Old 02-12-2016, 10:00 AM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
Quote:
Originally Posted by Bigus View Post
Driving resonance above the passband is a pretty good idea for sub enclosures (often not practical for fullrange cabinets), as it moves the most problematic frequency to a range where there is little energy left to excite it. You're then left with dealing with non-resonant vibrations.
Thanks for all the details, that was interesting and I think I have it now. The only comment you made I am still confused about is from your earlier post shown above. If we get the resonance above the full range speaker's passband I would think that would be a good thing as well, actually get it above human ability to hear? I guess I am confused when you said "little energy left to excite it?"

-Bob
bob_m10 is offline  
post #134 of 337 Old 02-12-2016, 10:38 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
Quote:
Originally Posted by Bigus View Post
Solid (volumetric) elements are the wrong approach for this sort of problem. A combination of plate elements and beam elements are perfectly adequate and will actually allow for a greater degree of resolution given a set amount of CPU resources and time.
I know, but I have a use in mind. I've made the bracing scheme much more efficient, but the middle 20% square inch area in two out of six faces are always inefficient to brace in such a manner (well, inefficient in man-hours).
ronny31 is offline  
post #135 of 337 Old 02-12-2016, 10:58 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5768
Quote:
Originally Posted by bob_m10 View Post
If we get the resonance above the full range speaker's passband I would think that would be a good thing as well
It's not a bad thing, but if the panel doesn't vibrate it doesn't vibrate, no matter what its resonant frequency. By and large when you add bracing the resonant frequency goes up, when you add mass it goes down. Either will give the desired result when the panel is either too stiff or too heavy for the internal pressures to cause it to vibrate.
Bill Fitzmaurice is offline  
post #136 of 337 Old 02-12-2016, 12:12 PM
AVS Forum Special Member
 
Bigus's Avatar
 
Join Date: Feb 2003
Location: The South
Posts: 5,560
Mentioned: 13 Post(s)
Tagged: 0 Thread(s)
Quoted: 587 Post(s)
Liked: 540
Quote:
Originally Posted by bob_m10 View Post
Thanks for all the details, that was interesting and I think I have it now. The only comment you made I am still confused about is from your earlier post shown above. If we get the resonance above the full range speaker's passband I would think that would be a good thing as well, actually get it above human ability to hear? I guess I am confused when you said "little energy left to excite it?"

-Bob
I sort of mixed more than one concept together there and wasn't very clear about it, so let me separate them.

It can be difficult to move the resonant frequency of an enclosure wall above the passband for a fullrange system, simply because 20kHz is a pretty damned high frequency! It would be great if you could, but it's generally impractical. Steel has a modulus of about 30 MPsi, and perhaps you could get there. For comparison, granite has a modulus of about 6 MPsi, fiberglass composite about 2.5, plywood 1-1.5, and MDF around 1/2 MPsi. So getting resonance that high with MDF is probably a non-starter. For full-range cabinets you are left with resonance somewhere in the passband (usually at many frequencies as you have panels and braces of various sizes), and have to deal with that accordingly. Damping plays a more critical role since it works particularly well at resonance, and you have to accept that you're going to excite panels at resonance from time to time. Stiffness and mass still work too, so use all techniques liberally (stiffness, mass, damping)!

You can move resonance out of the passband with sub enclosures (because you only need to push resonance up to a couple hundred Hz or so) fairly easily. There's little energy left to excite resonance in a sub enclosure with panels resonating at a couple hundred Hz, because it's far enough above the XO that you only have fundamentals down dozens of dB because of the XO as well as distortion products hopefully dozens of dB below the fundamental in level. So "little energy" at it's resonant frequency.

Of course panels vibrate at non-resonant frequencies as well, they just don't ring at those frequencies when the input goes away. Making something stiff and/or massive requires more input energy to produce a certain amplitude vibration, or alternatively a lower amplitude vibration for a given amount of input energy.
bob_m10 likes this.

Bigus is offline  
post #137 of 337 Old 02-14-2016, 05:42 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
Sitrep from the bracing simulation thread I made:


If you calculate the volume of the bracing in both instances you find that the corner bracing uses half the amount of material. There's a tad of optimization to do to get even better results with both simulations by placing them exactly 1/3 apart as the braces are a couple millimeters extra towards the center.
One among many findings in this new thread.
ronny31 is offline  
post #138 of 337 Old 02-14-2016, 09:26 PM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
So I have I have read and hopefully learned something about bracing thanks to the great discussion above. For my project, I am still thinking about using dowels but if you had to pick between the two styles shown in the attached pics which would you like better 1 or 2? the finished sub will have an outside dimension of 15.5 x 15.5 by 22.5".

Thanks
Attached Thumbnails
Click image for larger version

Name:	Brace v1.PNG
Views:	78
Size:	683.3 KB
ID:	1253745   Click image for larger version

Name:	Brace 2.PNG
Views:	61
Size:	118.9 KB
ID:	1253753  

Last edited by bob_m10; 02-14-2016 at 09:45 PM.
bob_m10 is offline  
post #139 of 337 Old 02-15-2016, 05:52 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
Quote:
Originally Posted by bob_m10 View Post
So I have I have read and hopefully learned something about bracing thanks to the great discussion above. For my project, I am still thinking about using dowels but if you had to pick between the two styles shown in the attached pics which would you like better 1 or 2? the finished sub will have an outside dimension of 15.5 x 15.5 by 22.5".

Thanks
https://www.avsforum.com/forum/155-di...l#post41574409

Cut a sheet with 45 degree ends and cut holes in it and install it like you would a normal brace we use everywhere. Placed so that they contact 1/3 the distance out on each sheet.

Or the more efficient designs of this corner-brace. The right one below:


Or more efficient bracing, instead of the right one above we make this brace, which uses even less material:

Only several of them. Placed so that there's an even distance between each brace contact point.
ronny31 is offline  
post #140 of 337 Old 02-15-2016, 06:01 AM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
Quote:
Originally Posted by Bigus View Post
Of course panels vibrate at non-resonant frequencies as well, they just don't ring at those frequencies when the input goes away. Making something stiff and/or massive requires more input energy to produce a certain amplitude vibration, or alternatively a lower amplitude vibration for a given amount of input energy.
Just a few more questions please, this is very interesting.

Ok so a light bulb just turned on, I was missing this important point. Moving resonance outside the speaker’s passband means the driver will have trouble getting the panel to vibrate since the fundamental will be below its resonance. It may still vibrate but once the signal stops it will not continue on its own. Hope I have that right.

You mentioned that damping is effective with resonance. How does gluing an inch of soft materiel on the inside of a panel wall stop resonance? Is it absorbing the waves that are bouncing around the inside of the cabinet?

Since typical wavelengths involved in the subs passband are really long and could not fit inside a sealed cabinet, what happens to them?

Why do sonotube subs not require bracing, is it because of the cylindrical shape or the material?

-Bob
bob_m10 is offline  
post #141 of 337 Old 02-15-2016, 06:14 AM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
Quote:
Originally Posted by ronny31 View Post
https://www.avsforum.com/forum/155-di...l#post41574409

Cut a sheet with 45 degree ends and cut holes in it and install it like you would a normal brace we use everywhere. Placed so that they contact 1/3 the distance out on each sheet.

Or the more efficient designs of this corner-brace. The right one below:


Or more efficient bracing, instead of the right one above we make this brace, which uses even less material:

Only several of them. Placed so that there's an even distance between each brace contact point.
Thanks, I wonder if it will be difficult to brace this way when you have a large hole for the driver in the bottom and another hole is the side for the amp. The rib or shelf design seem to work best in my box. I like the rib style so far because at least it touches all six sides.
bob_m10 is offline  
post #142 of 337 Old 02-15-2016, 06:36 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
This method of bracing is so effective that you need less than a quarter the material to get the same strength. Of course the way that uses FOUR times the amount of material is going to look "beefier", but it isn't.
The less material you have internally for bracing the more volume you have for added efficiency which means more db output at lower frequencies.

"it touches all six sides"... I do hope that you realize you would have this on all sides not just like what is in this illustration? This illustrates what each brace looks like, not how those braces are distributed throughout an entire enclosure.
ronny31 is offline  
post #143 of 337 Old 02-15-2016, 06:42 AM
 
Bill Fitzmaurice's Avatar
 
Join Date: Nov 2009
Location: New Hampshire
Posts: 15,491
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 3 Post(s)
Liked: 5768
Quote:
Originally Posted by bob_m10 View Post
which would you like better 1 or 2? the finished sub will have an outside dimension of 15.5 x 15.5 by 22.5".
1. 2 only braces in two dimensions.
Quote:
Why do sonotube subs not require bracing
Says who? A cylindrical shape gives stiffer walls for a given wall thickness, but that doesn't mean bracing may not be required.
Quote:
Moving resonance outside the speaker’s passband means the driver will have trouble getting the panel to vibrate since the fundamental will be below its resonance.
Not quite. If the panel resonance is within the pass band it will be easier for it to vibrate. But with sufficient internal pressure a panel will vibrate, no matter what its resonant frequency, unless its sufficiently stiff and/or has sufficient mass to resist the applied force. Conversely, with sufficient stiffness and/or mass a panel will not vibrate, no matter what its resonant frequency may be.
bob_m10 likes this.
Bill Fitzmaurice is offline  
post #144 of 337 Old 02-15-2016, 07:16 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
Here's an illustration of the fork shape bracing. This is however a slightly suboptimal distribution compared to the mathematical possibility from this brace format:

The corners can be more efficient using less material if the fork is redesigned for the corners. Whereas the middle forks are very efficient.
You can ofcourse use a simpler version that uses a lot more material (and gets a tiny bit more strength since it uses so much more material that you simply don't cut out from the forks):

But if you want low frequency tuning every liter you save in bracing is more db output.

PS: I haven't braced the side you are looking through.
ronny31 is offline  
post #145 of 337 Old 02-15-2016, 07:22 AM
AVS Forum Special Member
 
Augerhandle's Avatar
 
Join Date: Jan 2009
Location: About 25" away from my computer screen
Posts: 5,136
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 1150 Post(s)
Liked: 1140
Quote:
Originally Posted by ronny31 View Post
This method of bracing is so effective that you need less than a quarter the material to get the same strength. Of course the way that uses FOUR times the amount of material is going to look "beefier", but it isn't.
The less material you have internally for bracing the more volume you have for added efficiency which means more db output at lower frequencies...

I call bull*****. Stick bracing for a typical Marty Sub takes up less than 2.25% of the net volume. reducing that percentage will not make any appreciable difference to how the subwoofer works or sounds. One might as well install mini I-beams, it would still be as pointless.

"The wise understand by themselves; fools follow the reports of others"-Tibetan Proverb
_____________________ http://www.scientificamerican.com/article/auger-handle/ ________________________
Augerhandle is offline  
post #146 of 337 Old 02-15-2016, 07:31 AM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
Quote:
Originally Posted by ronny31 View Post
"it touches all six sides"... I do hope that you realize you would have this on all sides not just like what is in this illustration? This illustrates what each brace looks like, not how those braces are distributed throughout an entire enclosure.
Yes I do thanks.
bob_m10 is offline  
post #147 of 337 Old 02-15-2016, 07:42 AM - Thread Starter
Senior Member
 
bob_m10's Avatar
 
Join Date: Mar 2004
Posts: 330
Mentioned: 0 Post(s)
Tagged: 0 Thread(s)
Quoted: 88 Post(s)
Liked: 19
Quote:
Originally Posted by ronny31 View Post
But if you want low frequency tuning every liter you save in bracing is more db output.

PS: I haven't braced the side you are looking through.
It seems strange to me that you would use the efficiency of bracing to determine or increase DB output of the sub. I would imagine you would determine an appropriate enclosure volume during design phase of the sub and then look to efficient bracing schemes to shrink the size/weight or lower cost of building the box like Bill has been recommending.
bob_m10 is offline  
post #148 of 337 Old 02-15-2016, 07:51 AM
AVS Forum Special Member
 
Augerhandle's Avatar
 
Join Date: Jan 2009
Location: About 25" away from my computer screen
Posts: 5,136
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 1150 Post(s)
Liked: 1140
Quote:
Originally Posted by bob_m10 View Post
It seems strange to me that you would use the efficiency of bracing to determine or increase DB output of the sub. I would imagine you would determine an appropriate enclosure volume during design phase of the sub and then look to efficient bracing schemes to shrink the size/weight or lower cost of building the box like Bill has been recommending.

The difference is that Bill is actually qualified to design subwoofers, whereas Ronny31 is just spit-balling layman ideas.

"The wise understand by themselves; fools follow the reports of others"-Tibetan Proverb
_____________________ http://www.scientificamerican.com/article/auger-handle/ ________________________
Augerhandle is offline  
post #149 of 337 Old 02-15-2016, 08:13 AM
AVS Forum Special Member
 
Join Date: Mar 2015
Location: Norway
Posts: 1,170
Mentioned: 15 Post(s)
Tagged: 0 Thread(s)
Quoted: 675 Post(s)
Liked: 134
Quote:
Originally Posted by bob_m10 View Post
It seems strange to me that you would use the efficiency of bracing to determine or increase DB output of the sub. I would imagine you would determine an appropriate enclosure volume during design phase of the sub and then look to efficient bracing schemes to shrink the size/weight or lower cost of building the box like Bill has been recommending.
Sometimes the sheet material decides what is efficient. Like my white speakers:




The depth is exactly the max you can get from using half the width of the sheet material I used. And the height is exactly the max you can get from the height of the sheet material, if you want two sections in the height of the sheet. And the baffle is exactly the smallest size you can have, and still fit the woofer dimension by width. And just a few inches taller than it needs to be to fit the woofer, horn and two vents on the baffle face (so that the vents don't have to be elsewhere). The vent is precisely as long as is possible without having bends (which wouldn't fit because of the massive woofer basket). And the vent is precisely as big area as to avoid port noise (there's danger of port noise at 50-60hz at extreme output though, still haven't gotten the opportunity to test if the vents need to be slightly bigger for max output).
I built these two speakers from three 600x2400x22mm sheets of HDP and still had strips left for strengthening (but didn't use any strengthening, its so stiff material and wanted to know how exactly they would behave alone, these aren't the "production-run" versions after all).
Therefore, the response leaves something to be desired.

If I can make the next version of this enclosure stiffer than this one, yet have more internal room, even by a couple liters, then the output can be significantly better at low frequencies. Well, "significantly" by my use of the word. And just touching the faces, even the top and bottom faces that are reinforced with steel handles and speakon terminals, there's significant movement. Well, "significant" as I use it, I want zero detectable movement if I can get it. Then I can remove almost all the filler (there's about 2cm layer of fiber on all six surfaces).
I'm even trying to find out how much benefit there would be in changing from the 40kg/m^3 pillow-filler to 800kg/m^3 of coconut coir matting for erosion control.

You could of course use the more volume-efficient bracing to double or triple the amount of bracing you have. Perhaps not a benefit on certain subwoofers, but at least for LCR's its almost never enough stiffening (the way high end speakers get around this is simply to fill it with several layers of different density of stuffing material. Completely 100% stuffed speakers aren't uncommon). If I didn't want to lose bass output from my standalone LR setup I might just have filled them completely with stuffing as well.

Quote:
Originally Posted by Augerhandle View Post
I call bull*****. Stick bracing for a typical Marty Sub takes up less than 2.25% of the net volume. reducing that percentage will not make any appreciable difference to how the subwoofer works or sounds. One might as well install mini I-beams, it would still be as pointless.
If you have a certain external size limit, like me as explained above, then the more volume your bracing uses the less efficiency you get. For people who are so far along the line so as to have 4+ big subwoofers and whatnot, then 1db here or there is a lot of money. Doubling the amount of woofers and amps would add only 6db, so you want every bit of efficiency you can reasonably expect to get from what is very very very cheap comparatively (bracing the enclosure, even if complex bracing, its still cheap, it practically only costs time).
Btw, a typical marty isn't braced a lot. It doesn't really need to be. But in any case you could get a lot more output if you didn't need any filler in a marty because of having a very rigid enclosure. Well, "a lot more" being comparative to its cost. 1db more output from a double subwoofer setup by bracing and removing filler isn't difficult. If it costs you 2 grand to double your wooferage to 4 with twice the amps, then 1db for 10 bucks or something in glue in addition to some cut-off material, is a quite rational expenditure of calories.
If you get the material to reach farther, getting more stiffness for the same volume of material, its even better. And once we figure out a very very efficient bracing scheme, it will be as easy for everyone to copy as a marty design for UM18-22 or something.

Last edited by ronny31; 02-15-2016 at 08:19 AM.
ronny31 is offline  
post #150 of 337 Old 02-15-2016, 07:11 PM
AVS Forum Special Member
 
Augerhandle's Avatar
 
Join Date: Jan 2009
Location: About 25" away from my computer screen
Posts: 5,136
Mentioned: 27 Post(s)
Tagged: 0 Thread(s)
Quoted: 1150 Post(s)
Liked: 1140
Quote:
Originally Posted by ronny31 View Post
This method of bracing is so effective that you need less than a quarter the material to get the same strength. Of course the way that uses FOUR times the amount of material is going to look "beefier", but it isn't.
The less material you have internally for bracing the more volume you have for added efficiency which means more db output at lower frequencies...

Quote:
Originally Posted by Augerhandle View Post
I call bull*****. Stick bracing for a typical Marty Sub takes up less than 2.25% of the net volume. reducing that percentage will not make any appreciable difference to how the subwoofer works or sounds. One might as well install mini I-beams, it would still be as pointless.
Quote:
Originally Posted by ronny31 View Post

If you have a certain external size limit, like me as explained above, then the more volume your bracing uses the less efficiency you get. For people who are so far along the line so as to have 4+ big subwoofers and whatnot, then 1db here or there is a lot of money. Doubling the amount of woofers and amps would add only 6db, so you want every bit of efficiency you can reasonably expect to get from what is very very very cheap comparatively (bracing the enclosure, even if complex bracing, its still cheap, it practically only costs time).
Btw, a typical marty isn't braced a lot. It doesn't really need to be. But in any case you could get a lot more output if you didn't need any filler in a marty because of having a very rigid enclosure. Well, "a lot more" being comparative to its cost. 1db more output from a double subwoofer setup by bracing and removing filler isn't difficult. If it costs you 2 grand to double your wooferage to 4 with twice the amps, then 1db for 10 bucks or something in glue in addition to some cut-off material, is a quite rational expenditure of calories.
If you get the material to reach farther, getting more stiffness for the same volume of material, its even better. And once we figure out a very very efficient bracing scheme, it will be as easy for everyone to copy as a marty design for UM18-22 or something.
The problem is, that what you are claiming isn't remotely true concerning bracing. Cutting down on the stick bracing in a MiniMarty (or any) sub changes the volume insignificantly, and changes the tune even less. Where did you come up with 1 dB more output? Have you modeled it? I have.

Here's a screenshot of the WinIsd model of the MiniMarty. The red line is with standard stick bracing volume, and the black line is using only 1/4 of the bracing volume (Hint: The black line is behind the red line).






The change in bracing amounts to a drop in tune of about 0.15 Hz, but the peak output is louder (by .06 dB) with the standard bracing. Tell us all again how you can hear that, how it magically becomes 1 dB more instead of .06 dB less, or how else it makes a difference to the average DIYer.
Attached Thumbnails
Click image for larger version

Name:	Capture MiniMarty Bracing.JPG
Views:	250
Size:	135.7 KB
ID:	1256865  

"The wise understand by themselves; fools follow the reports of others"-Tibetan Proverb
_____________________ http://www.scientificamerican.com/article/auger-handle/ ________________________

Last edited by Augerhandle; 02-15-2016 at 07:56 PM.
Augerhandle is offline  
Sponsored Links
Advertisement
 
Reply DIY Speakers and Subs

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page


Forum Jump: 

Posting Rules  
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off