| m0tion |
I was reading an article describing how the Linkwitz Transform circuit works at ESP and I saw something that really confused me:
"Bracing is usually not needed, since the frequencies are so low that panel resonance is unlikely if the unit is a self contained subwoofer."
You can read the article for yourself here. I've always heard bracing was VERY important to speaker enclosures ESPECIALLY subwoofers, is this wrong? I'm about to start construction on a sub, but I don't want to waste my time on bracing if it isn't important. Anyone else have insight they'd like to share? |
|
|
| richie00boy |
What he means is that if you use a Linkwitz Transform or EAS circuit then the chances are very high indeed that you will be using a small box. A small box does not have the large panel areas to vibrate and are inherently more rigid anyway, the result being that bracing is not always strictly necessary.
Consequently, if your box was larger than about 30 litres, bracing would be required. |
|
|
| sek |
Hi,
in general, there are two reasons for bracing an enclosure, both have to do with damping. While the first use for braces is damping the standing waves inside the enclosure, the second reason for braces is damping the resonating enclosure walls.
For subwoofer (or bass box) applications, damping the internal box volume is useless, while stiffening the box enclosure is essential.
Both damping purposes are done by bracing the box internally, but the braces for the two purposes differ in the way they are constructed.
I agree that a very small box volume (made of reasonably stiff material) doesn't really need bracing. It would ruin the effect of having the smallest possible enclosure anyway...
Sebastian. |
|
|
| Yury |
| quote: | Originally posted by sek
Hi,
I agree that a very small box volume (made of reasonably stiff material) doesn't really need bracing. It would ruin the effect of having the smallest possible enclosure anyway...
Sebastian. |
that's right about small box, but anyway wood will be resound.
that's because I've use material with large number of damping factor. it's to answer the purpose of first reason.
second one - standing waves - very serios problem and we can't leave it out, using bracing.
form of box, damping materials inside...should be effective
ok. good luck |
|
|
| m0tion |
| quote: | | For subwoofer (or bass box) applications, damping the internal box volume is useless, while stiffening the box enclosure is essential. |
Could you clairify this statement? By "damping the internal box volume" do you mean using damping material to line the walls and stuff the box? I was using the impression this was useful. The enclosure I will be using is about 3.7ft^3, pretty large, I believe bracing would be appropriate then? |
|
|
| tiroth |
The usual damping materials (bitumised felt, etc) aren't really effective at sub frequencies. In that sense, damping is not that useful--the fill is hopefully enough to absorb higher frequency low amplitude harmonics, suspension noise, etc.
Open cell foam is a little different.
Extreme mass loading, such as concrete or other heavy lining, is probably in a different ballpark though. |
|
|
| BillFitzmaurice |
Bracing is critical for a sub, more so than other speakers due to the high internal pressures involved. A good rule of thumb is to have all panels fully shelf braced so that at least one plane has no more than 8 inches of clear span. That pretty much limits maximum box size without braced panels at 1/3 cu. ft or 30 liters.
Internal damping is only required when any internal dimension measures at least 1/4 wavelength; for a sub with a 100 Hz low pass function that translates to about 2.8 feet.
www.billfitzmaurice.com |
|
|
| sek |
| quote: | | Could you clairify this statement? |
Damping standing waves or resonances inside the box is usually done with soft, absorbant materials (to line the walls and stuff the box). I just wanted to say that this is almost useless in bass applications, as the damping effect of e.g. foam, cotton or other soft materials is frequency dependent - and due to the long wave lengths of bass frequencies, the structure of the soft damping materials doesn't absorb much energy (because the tissue fibers or foam bubbles don't "resonate" with the waves).
Damping with soft, absorbant materials comes into play from the mid-bass upwards and has it's strongest effect from the upper-mid frequencies on.
| quote: | | The enclosure I will be using is about 3.7ft^3, pretty large, I believe bracing would be appropriate then? |
That's pretty large, right. You should apply bracing to the walls as Bill recommends. A beautiful method is to build a matrix of boards so that the boards develop "tiny" cubic volumes, which are cut out. This would then result in a 3D-frame structure filling the entire box volume (don't forget that the speaker opening needs some spare volume for the speaker back side and magnet).
This gives the two advantages that the outer box walls are covered with stabilizing elements while the pairs of opposite walls are stiffly connected. Thus, the pressure can't expand or collapse (simultaneously push out or in) the enclosure walls. The smallest possible matrix size would of course be 2, one big, two medium and four small boards cut out and glued in... I usually make them out of thin MDF, as it's easy to work with and you don't need anything more strong here. But stick with Bill, your box size probably needs more than one brace per wall dimension. What board material and thickness do you use?
;)
Hope this helps,
Sebastian. |
|
|
| Kittle |
okay im confused now too....
bracing? no bracing?
damping on the walls? bare wood?
stuffing? no stuffing?
I just finished a sub of my own - 3.5 cubic feet. The entire insides of the box is lined with BH5 and there are no braces..
but it sounds great!
Standing waves? .... seeing as there are no internal braces im assuming i'll get them. What do they sound like?
Stuffing a box always seemed counter productive to me.. we work so hard to get the internal volume correct (subtract port size, subtract speaker cone displacement, subtract bracing, etc...) - and then said volume gets reduced by some arbitrary amount of stuffing.
someone edumacate us please! |
|
|
| m0tion |
| quote: | | Stuffing a box always seemed counter productive to me.. we work so hard to get the internal volume correct (subtract port size, subtract speaker cone displacement, subtract bracing, etc...) - and then said volume gets reduced by some arbitrary amount of stuffing. |
The design I'm discussing is a sealed box design, a lot of these suggestions may not apply to your ported (bass reflex) sub. Just wanted to make sure to reiterate that. Also, stuffing actually increases the apparent box volume by causing air to move more slowly throughout your ported enclosure. |
|
|
| paulspencer |
Motion,
great website! Nice to see someone who does just about everything diy :D ... would be great to see a write up on the amp and speakers as well.
...
1/3 cu ft = 30L :xeye:
>>> That should be 1 cu ft = 28L
...
I don't see how standing waves have any relevance inside any speaker box.
The rear wave that is trapped inside the box will do the following:
1. be absorbed ie. converted into heat inside the box
2. re-enter the room via the speaker cone or passive radiator (the driver is a weak part in the system as far as transmission through the enclosure is concerned)
3. re-enter the room via a port
4. cause vibration of the box and thus be re-radiated into the room with a change in frequency response due to the stiffness and damping of the box and resonances in the transmission characteristics of the box ... thus this involves "coloration"
As damping is not effective below 100 Hz, the best thing you can do is make the box as stiff as you can. The walls should be reasonably thick and well braced as Bill mentioned. Another technique not commonly used is called constrained layer damping (CLD), which is in effect building a box within a box. The driver will be attached to the inner box and the two boxes will be connected via a flexible adhesive like liquid nails.
I know it has been said that bracing or internal baffles to deal with standing waves is not for small subs. However, I don't see how standing waves have any real effect on box coloration in any box.
The question of bracing and box thickness - there is a point where you keep making your box thicker and adding more bracing where the box just becomes bigger and there is no noticeable improvement. |
|
|
| marsupialx |
A few weeks ago, I wanted to see what sitting on a small commercial sub would feel like. We played the explosions from U571 (the movie). The vibrations were so intense it was almost painful, I had to clamp my jaw to keep my teeth from chattering.
My friend said the apparent bass in the room was reduced by about 10dB. This ponderous smeared cabinet resonance was a major contributor to the sub's output.
My subs at home are built with a box inside a box, with damping material in between, and very well braced. You can not feel any vibration on the exterior box.
Damp, brace, mass load, and spike-couple them to the floor. You will get clean non-smeared bass as a reward. |
|
|
| sek |
| quote: | | I don't see how standing waves have any real effect on box coloration in any box. |
A standing wave is essentially a resonance with a high Q (low damping factor). As all parts in the box interact (acoustically), the Q of the box resonances have influence on the response of the speaker in it's box.
A completely undamped standing wave inside the box will - once energized - have an easily perceptible influence on how the speaker's cone movement at that very frequency decays. It just takes longer for the cone to stop, once the wave is "standing". ;)
That's why damped boxes sound less "boomey" than before damping them... |
|
|
| BillFitzmaurice |
| quote: | | I don't see how standing waves have any real effect on box coloration in any box |
They don't if the box is too small for them to develop. In the pre- T/S days cabs with at least one internal dimension of 3 feet or more were common and standing waves could be real problematic, as they will occur when that distance is 1/4 wavelength or more, and at 3 feet that means about 100 Hz. It took a lot of fiberglas or wool or whatever to damp out the resulting resonant peaks and valleys at 100 Hz, or even 200 for that matter. Nowadays with smaller box sizes standing waves are seldom an issue; the midrange frequencies where they can develop are quite easily damped with an inch or so of polyfill batting on the box interior walls.
Don't confuse damping with vibration control. Vibration control devices, be they braces or asphalt sheets, control physical vibration of the cabinet walls, while damping materials control wave reflections inside the box.
Sorry about the English/Metric comversion. I still prefer Henry VIII to Napoleon. |
|
|
| marsupialx |
The dictionary definition of dampen. Seems like it would apply to cabinet walls as easily as to air.
'to check or diminish the activity or vigor of : DEADEN'
However, there may be an agreed upon limitation of its usage in audio engineering that I am not aware of. |
|
|
| m0tion |
| quote: | | My subs at home are built with a box inside a box, with damping material in between, and very well braced. You can not feel any vibration on the exterior box. |
I haven't heard a lot about this technique (probably because it involves a LOT more work than a traditional cabinet). I'd definately be interested in hearing the details of this construction and pics would be great as well. I like the idea though and I'll give it some serious thought. If nothing else could you give a brief description of your particular implementation? |
|
|
| paulspencer |
I've made an animation showing the construction of a sub using CLD (a box within a box). The driver is attached to the inner box which is braced using matrix bracing in 3 planes. The vibration of the driver will transfer to the inner box, and this will be damped by the flexible adhesive layer before being transferred to the outer box, with a magnitude that is significantly reduced.
Shown in the animation:
450 x 450 x 450mm sealed box
18mm MDF
driver: Stryke AV12
Box volume: ~50L |
|
|
| paulspencer |
| Looks like it didn't come through ... 2nd try: |
|
|
| paulspencer |
| quote: | Originally posted by marsupialx
Damp, brace, mass load, and spike-couple them to the floor. You will get clean non-smeared bass as a reward. |
I can't see any acoustic virtue in spike coupling a sub to the floor. What this will do is transfer vibrations through to the floor, and if the floor is not very solid (such as concrete), but is in fact timber, then the vibrations will cause the floor to make a contribution to the sound that will be less true to the original as it will add its own resonances. This *might* result in a sound that you like more, but as far as accuracy goes it works against the other measures you have taken. To be consistent, I think you would be better to use a flexible coupling to the floor which reduces the transfer of vibration.
| quote: | Originally posted by sek
A completely undamped standing wave inside the box will - once energized - have an easily perceptible influence on how the speaker's cone movement at that very frequency decays. It just takes longer for the cone to stop, once the wave is "standing". ;)
|
Sek, based on your description, I'm not sure if this is true or not. Do you have any links to other information to back this up?
| quote: | Originally posted by BillFitzmaurice
Don't confuse damping with vibration control. Vibration control devices, be they braces or asphalt sheets, control physical vibration of the cabinet walls, while damping materials control wave reflections inside the box.
Sorry about the English/Metric comversion. I still prefer Henry VIII to Napoleon. |
very good point
...
As I'm in Australia (metric), and there are so many people from the US in audio discussions and with websites, I have to be good at metric/imperial conversion. Generally, I find imperial units a nuisance, but when it comes to audio, often imperial are more intuitive - it's easy to visualise a 1 cu ft sub box or the size of a 6.5" driver, etc.
| quote: | Originally posted by m0tion
I haven't heard a lot about this technique (probably because it involves a LOT more work than a traditional cabinet). |
Motion, perhaps you are overestimating the amount of extra work. If you use a flexible adhesive, the main difficulty is getting the gap between the boxes right - too small and you won't get enough adhesive in, too big and you won't get sufficient contact for the adhesive to work properly. It's not a lot of work to cut a second box, it's just a matter of being more precise. |
|
|
| sek |
Hi Paul,
| quote: | | Sek, based on your description, I'm not sure if this is true or not. |
I could explain myself a little better. ;)
A standing wave (room resonance) takes it's time to develop. It doesn't "swing in" instantaneously but builds up energy over at least one period of cone movement. One half for the first half wave to travel to the opposite wall, another half wave to travel back and then add constructively to the next pressure change. Correct me on that if I'm wrong.
Now that's stored energy. Whatever it takes to stop this standing wave from bouncing back and forth between the opposite walls (e.g. damping material), it won't stop in no time. The stored energy has to go somewhere. Damping material is there to absorb this energy (and thus probably prevent the standing wave from occuring in the first place) and convert it into heat.
But let's assume there is no damping at all (an impossible condition in reality). What should the stored energy do, "stand" forever? No, because there is the speaker cone. Model-wise, both the vibrating air of the standing wave and the cone-motor-spider-suspension system can be seen as mass-spring systems. One has stored (kinetic) energy while the other (the speaker) may or may have not.
For simplicity, let's assume the speaker cone is completely stopped while the standing wave is still bouncing. Now those two mass-spring system are directly coupled, as the air is directly touching the speaker. They must interact, except the speaker had infinite damping (e.g. an ideally stiff cone and suspension), which it definitely hasn't, as it is a speaker. ;)
Why should it be possible that the energy in this directly coupled mass-spring system can travel in one direction while it can't in the other? At least, the speaker has built up the energy stored in the standing wave...
In reality, this process is heavily depending on the damping properties of the box and the mechanical and electrical properties of the speaker. The process is extremely frequency dependent and it's likely that you never experience such problems by just building good boxes :D
| quote: | | Do you have any links to other information to back this up? |
No, sorry. That is from (german) textbooks and personal experience (experiments) in class and in DIY. And it could be wrong, don't forget that ;D (everybody jumping in, please)!
But as I said, the difference in "reverberance" between an undamped box and the same box filled with damping material is easily audible and shows up as delayed decay at certain frequency bands (remember, the speaker's properties come into play here, different speakers of the same size would decay differently, time- and freuqency-wise).
puuh, 'nuff. ;)
Sebastian. |
|
|
| m0tion |
| quote: | | The vibration of the driver will transfer to the inner box, and this will be damped by the flexible adhesive layer before being transferred to the outer box, with a magnitude that is significantly reduced. |
Cool animation, but I'm still a little fuzzy on this point. The driver is mounted to the inner box, but it still has to be exposed to the room, you you'll need a hole for it in the outer box as well right? What does the transition from inner box to outer box look like, or, how is it accomplished? Doesn't this also cause the woofer to be severly recessed into the box? Is that not a problem? |
|
|
| paulspencer |
Here's a plan to describe (actually it's a section).
The driver is attached to the inner box (magenta), but not flush mounted as you would normally do. The red shows the bracing.
The yellow is the outer box. The hole is cut larger to accomodate the outer diameter of the face plate.
The dotted blue line (ignore the little squares in the corners) is the damping layer, which would be the liquid nails adhesive.
No, as you can see, the woofer is flush mounted as you would normally have if you had double thickness walls, as some do. The difference is that the adhesive is flexible and the layers are not rigidly connected.
Basically you build the bracing, then build the inner box around it. You then partly assemble the outer box, apply the flexible adhesive damping layer, and progressively add the outer box, then fix the driver. It requires a little more care, but isn't by any means the most difficult sub box to build. |
|
|
| m0tion |
| Thanks for the drawing, I think I fully understand what you guys were talking about now, but I've got a question (or maybe a suggestion). I think the construction could be made much more simple if the front baffle did not have a layer of damping material between the inner and outer walls (see pic, sorry paul, I kind of mangled yours). Would this still provide much the same effect? I'm guess it would for the most part, but the more isolated the inner box is the better. |
|
|
| paulspencer |
I assume what you are talking about is removing the damping adhesive layer on the front baffle. This is probably not a good idea as the outer and inner box would then be connected on the front baffle and vibrations would transfer through. This would mean a loss in performance without any real simplification in construction. If you want to build a box within a box following this CLD technique, then you pretty much have to do it as I've shown, otherwise you might as well just make the walls thicker or use bracing as you see required. Otherwise you end up with the extra effort of CLD construction (although it isn't that bad) with performance that is no better than a more conventional solidly built box.
Now if I misunderstand your description and you in fact mean make the front baffle just one layer thick, then this becomes a significant weak point in the box, you certainly don't want this. |
|
|
| m0tion |
| Yes, I was talking about removing the damping adhesive layer on the front baffle. Another question for you. How did you decide on using liquid nails as a damping layer? Do they sell liquid nails in large amount (liter/gallons)? The first material that came to my mind would be to use neoprene foam rubber sheets (which can be purchased from www.mcmastercarr.com, search for "neoprene foam rubber"), what are your thoughts on this? Also, how much better does this technique work than simply lining the inside wall of a normal enclosure with liquid nails (or neoprene)? |
|
|
| paulspencer |
Liquid nails adhesive is recommended by Dr Earl Geddes
In suggesting liquid nails I'm effectively passing on his recommendation as it makes a lot of sense. Yes you can use foam but it won't have the same effect. There are 2 reasons to use liquid nails:
1. being an adhesive, the bending of the panels that takes place in vibration transfers shear forces to the adhesive, thereby maximisin the damping effect
2. vibration of the inner panel is not coupled to the outer, thus the magnitude of vibration is damped and reduced before being passed on to the second layer
The first action is often not understood, therefore foam is considered suitable. I believe a flexible adhesive is best, I'm sure there are others like Liquid nails that are just fine.
So you can see if you line the interior with liquid nails it will do virtually nothing. In Australia you can get liquid nails in cylinders similar to those for silicone sealant.
I think CLD is a great technique to use as it maximises effectiveness very efficiently - it will perform significantly better than a conventional box of the same thickness. Compared to a double thickness box, you only have to add adhesive and some extra attention and care in construction. |
|
|
| m0tion |
| Of course you're right. I see how the CLD technique could be very effective at mechanically isolating the inner box. I understand that the liquid nails would do a good job coupling the inner box and the outer box, but I think the same effect could be achieved by using a 1/2" thick sheet of neoprene contact cemented to both the inner and outer boxes without the need for large amount of adhesive (I use liquid nails on my crossovers and think it has a strong smell, I can only image what it would be like pouring it into a speaker enclosure). I'm now strongly considering this technique for my next subwoofer. Any other comments you have would be appreciated, thanks for taking the time to explain yourself. |
|
|
| paulspencer |
| quote: | Originally posted by m0tion
I understand that the liquid nails would do a good job coupling the inner box and the outer box, but I think the same effect could be achieved by using a 1/2" thick sheet of neoprene contact cemented to both the inner and outer boxes without the need for large amount of adhesive (I use liquid nails on my crossovers and think it has a strong smell, I can only image what it would be like pouring it into a speaker enclosure). |
;) Yes that would work fine. You could first cut and assemble the inner box. Then place the neoprene over it to get the right gap between the two layers. Then you might trim the outer box when placing the panels over the neoprene, having first made it slightly larger than required.
| quote: | Originally posted by m0tion
I'm now strongly considering this technique for my next subwoofer. Any other comments you have would be appreciated, thanks for taking the time to explain yourself. |
If you do, it would be great to see some pics on your site of construction ...
My subs are in prototype stage, but when I get around to building the final box, they will be constructed using 3 techniqes to get an inert box:
1. a fairly elaborate bracing mechanism
2. curved walls, mainly to get the look I want, but also allows me to make the walls a little thinner (as seen in the rendering on my site)
3. CLD - trickier to achieve with the box I have in mind as it incorporates 3mm sheets of MDF laminated together to form a curve |
|
|
| Christof |
Ok, I got my dead horse beater in hand and I'd like to get more info from this old thread. As an architectural woodworker I use this liquid nails adhesive on a a semi-regular basis. I know for a fact that this adhesive gets very hard (semi-rigid) after a period of time. I do not understand how it can effectively absorb energy and turn it to heat via friction as we are trying to acheive with CLD.
I'm in the process of building a new HTsub and would really like to hear what other materials are used for CLD construction?
I plan to press my 3 layers (1/2" MDF + 1/4"?? + 3/4" MDF)together in a vacuum press with some sort of layer between. All joinery will be done with mitres such that the three layers will be aligned to form a CLD enclosure.
Thanks
Christof |
|
|
| m0tion |
| Funny you should ask. I'm at this time preparing a webpage that will describe the steps I used to build a subwoofer I just recently finished. It used this so called "CLD" technique, except I used a 1/2" layer of neoprene foam rubber as the mechanical isolator. I'll be sure to make a thread once I finish the guide. I do already have a few working pictures posted in my "Current Theater Pics" section. |
|
|
| paulspencer |
| quote: | Originally posted by Christof
I know for a fact that this adhesive gets very hard (semi-rigid) after a period of time. I do not understand how it can effectively absorb energy and turn it to heat via friction as we are trying to acheive with CLD. |
It needs to remain flexible for this to work. I first heard liquid nails recommended for this by Dr Earl Geddes at http://www.gedlee.com/
Are you saying that you don't think it has enough elasticity? Or that it becomes rigid and brittle over time? |
|
|
| Ron E |
| quote: | Originally posted by Yury
that's right about small box, but anyway wood will be resound.
that's because I've use material with large number of damping factor. it's to answer the purpose of first reason.
second one - standing waves - very serios problem and we can't leave it out, using bracing.
form of box, damping materials inside...should be effective
ok. good luck |
Marble has a high damping factor? ;)
Take a piece of plywood, and a piece of marble. Hang them from one corner and strike them. Measure the time it takes for the sound to die out. If the marble is not cracked, it will ring like a bell. The wood won't. Feel them after 5 seconds, which is vibrating more?
Marble has high mass and stiffness, but damping is intrinsically bad in most stiff materials. Actually, one guy tested many materials for panel resonance (including stone IIRC) and found particle board (not MDF) had the best properties among easily available materials.
The walls in old buildings are really thick stone, not because that is what is required to resist attack or collapse, but because they had no idea how to decide how thick to make the walls.. Many DIYers are the same way - overbuilding because if 3/4" is good, 2" must be better......
Pressure changes in boxes are about 1Psi or 7kPa for large excursions in small boxes (when Vd = 0.05*Vb) |
|
|
| planet10 |
| quote: | Originally posted by BillFitzmaurice
vibration control. |
Push-push loading is a huge benefit here. I'm at the point where i usually don't consider a sub unless i have an even number of drivers for it.
dave |
|
|
| paulspencer |
| quote: | Originally posted by planet10
Push-push loading is a huge benefit here. I'm at the point where i usually don't consider a sub unless i have an even number of drivers for it.
|
Dave, I like this idea but I find it tends to mean smaller cheaper drivers, or big boxes! (If we are talking serious subs). It also makes it very difficult to consider a sub like the Tumult, although I suppose you could look at a pair of 15" Koda drivers using the same XBL2 technology. |
|
|
| planet10 |
| quote: | Originally posted by paulspencer
I like this idea but I find it tends to mean smaller cheaper drivers, or big boxes! |
I feel the advantages of push-push (particularily for a sub) give you a better end product.
One can see the effort you are going to to keep the outside of the box from transmiting speaker induced vibration, push-push kills it at the source (and at the same time improves downward dynamic range). My single 3/4" layer push-push box will have less external vibration than your double-layer box, and will have much less problems of energy storage (and later release) that you have to worry about in such a box
dave |
|
|
| Tim Moorman |
For a closed box system, I think room effects are so dominant at subwoofer frequencies that two separate subs in solid, well braced boxes will be of far greater benefit than the dual P/P or other single box configurations, no matter how well the box is built.
Dipole open baffle bass is a good solution for the room, as far as they go, perhaps with twin subs below for HT.
Tim |
|
|
| planet10 |
| quote: | Originally posted by Tim Moorman
For a closed box system, I think room effects are so dominant at subwoofer frequencies that two separate subs in solid, well braced boxes will be of far greater benefit than the dual P/P.
|
You underestimate the effectiveness of push-push loading... but i do agree 2 woofer boxes are better than 1... so my even number of choice is also evenly divisible by 4.
dave |
|
|
| jdybnis |
When you do a sub push-push, where do you put the drivers? Can you position it in a corner or are you restricted to putting it out in the open?
-Josh |
|
|
| planet10 |
| quote: | Originally posted by jdybnis
When you do a sub push-push, where do you put the drivers? Can you position it in a corner or are you restricted to putting it out in the open?
|
Drivers are placed on opposite sides of the cabinets (ie back & front or left side & right side -- or if lucky one driver on each side)... i have at least one client that uses his in a corner -- set on an angle
If you were designing for a corner, you could make the box 5-sided to have it stick into the room less (in this case the plate amp faces the corner)
dave |
|
|
| Svante |
I missed this thread. Anyway, I thought I should share my view on the original question, wether or not bracing is more or less important in subs.
If we assume that the lowest resonance in the walls of the box is higher than the upper frequency limit of the crossover, the motion of the walls will be spring-controlled and there will be no excitation of the resonance. This means that the walls will vibrate, and the displacement of the walls will be proportional to the pressure inside the box. The box will expand and contract (is that the proper word?) like a balloon. Stiffer walls will reduce the displacement of the walls.
Now, vibrating walls might seem like a bad thing and if they vibrate a lot, it is. However, since the frequency response of the walls is pretty flat (ie below all resonances), the vibration of the walls will not contribute a lot to the overall response of the box. Since there are no sharp peaks (resonances) in the box contribution, it will hardly be visible in the overall response of the system.
There is of course a limit to this reasoning. If the walls vibrate to such an extent that they become non-linear and produce distortion, this will be bad. So, a card-board box will not do, but that is probably obvious. Also, if the vibration is transferred to the floor or some other big surface that has resonances in the sub's frequency range, it is bad too. But I see little need for concerns about vibrating walls in a normal sub's frequency range, since there typically are no resonances in the cabinet there.
For full range systems, on the other hand, there will definitely be structural modes in the frequency range of the speaker, and the need to control the box vibration will therefore be larger. Since a resonance will yield a contribution in a narrow frequency range, the effect on the response will be more obvious. Sandwich damping can be used to reduce the Qs of the structural modes and bracing can be used to move the modes towards higher frequencies, where the excitation is less efficient.
Just my 5 öre... |
|
|
| Tim Moorman |
Yeah, those woofers do mount up. Great bass can get costly.
What drivers and box sizes are you dealing with in the p/p sub?
Tim |
|
|
| planet10 |
| quote: | Originally posted by Tim Moorman
What drivers and box sizes are you dealing with in the p/p sub?
|
My beater woofers are vintage 12" Fosters... they work well in a 2.5 ft^3 aperiodic box. I have 4 Peerless 8s (go lower but won't move as much air, that will likely go in a TL, and long-term i am working to earn 4 deluxe 10" to go in TLs.
dave |
|
|
| Tim Moorman |
Peerless does offer a good value for the buck. Have a look at the 850146 for a decent 10" at affordable price. For a medium size room they should be fine doubled up in p/p, or separate boxes.
I like stereo bass (as you've probably gathered). There is really much more "there" there than one would imagine. The soundscape grows fuller, and much more realistic. I wouldn't have thought this possible since bass below 100 Hz is supposedly omnidirectional. Not so, in my experience - the source is easily localized if the crossover is up near 100 Hz, and especially easy with a 12 dB (only) high pass filter is in use. Crossed low with 24 dB HP filters, and things get much harder.
The other problem that I can see with p/p is an application which requires high output due to room size, along with deep extension. Many of the true subwoofers are 4 ohm impedance only, and doubling these up requires a very substantial amplifier for the 2 or less ohm load, or the use of an isobaric pair each side for impedance matching. Ha! That might be interesting.
Tim |
|
|
| planet10 |
| quote: | Originally posted by Tim Moorman
Many of the true subwoofers are 4 ohm impedance only, and doubling these up requires a very substantial amplifier for the 2 or less ohm load |
Or wire them in series (i'll even do this for an 8 ohm speaker -- lower amplifier distortion) or what i usually do -- use 4 amp channels.
dave |
|
|
| SmarmyDog |
Another reason for bracing could be to keep the sealed cabinets' walls from physically puffing out and sucking in at sub-resonant frequencies. Granted the actual displacement could be quite small, perhaps 1mm or less, but if you factor in the surface area it seems it could start to steepen your low-end rolloff.
-Casey Walsh |
|
|
| Tim Moorman |
Planet10
4 channels? Actively, I presume?
I just got the MCM flier which shows the AudioSource AMP offerings with the AMP 300 capable of 235 watts each (2) channel into 4 ohm, is 2 ohm stable, has auto sensing, some adjustable gain limiting on the back, but level adjustment on the front, balance, a pass through line level feature on one of two inputs, and can be 375 bridged watts, 8ohm. $400. Also, $300 for the AMP200 at 125watts x 2, 200 w bridged, same features.
Smarmy
Not sure I'm following. Yes, a poorly built box can expand and contract minutely, I suppose. But more likely it will have a lot of spurious energy in the panels generating their own racket and muddying up the sound. It may sound bad, but it shouldn't effect roll-off. It might effect rolling out to the curb for pick up.
Tim |
|
|
| paulspencer |
| It would be interesting to construct a subwoofer in a box with two fully sealed enclosures for the front and rear of the driver, separated by a baffle. This would be a test of the impact of the sound from the box, as this would be all you would hear. |
|
|
| lopan |
[QUOTE]Originally posted by m0tion
[B]
Also, stuffing actually increases the apparent box volume by causing air to move more slowly throughout your ported enclosure.
Actually this is dependent upon the density of the stuffing. Most people don't stuff anywhere near what is required to generate a an apparent volume increase. Probably around 2.5 lbs per cubic foot may do this |
|
|
| Tim Moorman |
Huh? Try stuffing 2.5 lbs of fluffed poly dacron into a cu ft sometime. A one lb bag is about the size of a 30 gallon kitchen bag filled. You'd be lucky to get 2.5 oz.
Vance Dickason used a combination of fiberglass insulation and Acousta-stuf for best results.
Tim |
|
|
| lopan |
| I only work with long hair wool and cotton they can easily be stuffed to these density. Wouldn't want to work with fiberglass |
|
|
| creek |
| quote: | Originally posted by planet10
Drivers are placed on opposite sides of the cabinets (ie back & front or left side & right side -- or if lucky one driver on each side)... i have at least one client that uses his in a corner -- set on an angle
dave |
planet10, is that a brace attached to the back of the opposing drivers in your drawing? |
|
|
| BassAwdyO |
| I myself have built subwoofer enclosures using quite a few mountings. Push Push definately did get the ENTIRE box to vibrate less, as a whole. Panel resonance has nothing to do with push push mounting. Push push merely cancels out the intertia forces created by the drivers. If the speaker cone moves out, then it will push the box backwards (in the opposite direction of the cone movement). Its not causing a panel to vibrate, its causing the box as a whole to vibrate. I've built a push push design and used the same drivers that I formerly had mounted on the same side of a similar alignment. The push push design had a slightly higher Q but both enclosures sounded very similar. I do not think push push design gives much, if any, audible advantage. I have used upward firing woofers in sealed enclosures and found that it does make the floor and walls of the building structure resonate ALOT more. This might cause slight coloration, but what I really noticed was pictures rattling etc... It did give tremendously more impact to the sound however. Bass drum hits could be felt much more than they could before at the same level, explosions on movies etc... I'd say side firing woofers push push or not shouldnt have much problem with whole enclosure vibration. As for panel vibration 3/4 MDF shouldnt have any resonance below 100hz unless you're using long and moderately wide panels. I'd say bracing is needed at least every 12 - 15 inches. Bracing every 8 inches will work fine, but it might be slightly excessive. As for the Box in a Box design, its an interesting idea, but I'd bet if we did a listening test comparing the same driver in the same alignment crossing over below 100hz at 24db/octave no one could tell a difference. The extra time, and extra weight of the enclosure are particularly unattractive. My sub already weighs 130lbs with the driver. |
|
|
| Circlotron |
| Even if your panels don't hit a resonance, if your total panel area is large compared with the driver effective cone area it won't take very much flexing of the panels to almost equal (and therefore undo) the air displacement of the driver. The panel area of mine is about 52 times that of the 12" cone. |
|
|
| BassAwdyO |
"Even if your panels don't hit a resonance, if your total panel area is large compared with the driver effective cone area it won't take very much flexing of the panels to almost equal (and therefore undo) the air displacement of the driver. The panel area of mine is about 52 times that of the 12" cone."
-circlotron-
Have you tested this? I dont really have anything to back me up here, but I would imagine that panel vibration works somewhat like the vibration of a Passive Radiator because that is in fact what the panels are doing, passively radiating the sound. Anyhow, if your 12" sub movies even 2 inches p-p then your panels will have to move over a 32nd of an inch to cancel out the driver's acousitc output. Thats a serious underestimation considering not every square inch of surface on your box is going to move that much. Only the middle of the panels will move a significant amount, not the corners where the boards join. I'd say you'd be getting near 1/16 or 1/8 of an inch of movement. Unless your building your box from some cheesedick 4 ply 1/2 inch pine (which you should NEVER do) that much panel flexing should not occur. As I said before bracing every 12-15 inches should make your box solid. If you have a box less than 15x15x15 I'd say bracing is very optional. It would only be needed with a extremely powerful 12" driver, probably more powerful than those currently on the market. I'd put a few braces for something like a brahma, but I really dont think it would make a big difference. |
|
|
| Svante |
Hmm, I was thinking about how to estimate the actual movement of the box. How about this: Play some loud music. Feel the vibration using the fingers on the box. Then lower the volume, and put the fingers on the woofer. Raise the volume until the vibration feels equally strong. Check the level difference in between the two cases. Add 20*log(Sb/Sd) to compensate for the larger box area (Sb). I think there will be plenty of margin until the box palys as loud as the speaker, even with pretty thin walls.
Hmm, I'll have to test it.
Edit: I did it, I had to lower the level by some 50 dB using a 100 Hz sinusoid on my pretty weak boxes (18mm softwood, box is 0.2x0.2x1 m) without bracing. 50 dB corresponds to 316 times when it comes to surface, assuming that the entire surface moves like the spot where I put the fingers. I don't think it does. Sb is 61 times Sd on my system. Not very scientific, but it gives an idea about the order of magnitude of the vibs. |
|
|
| lopan |
| People seem to be talking about sub boxes as is they only resonate within the subs pass band (cabinet breathing). The arger issue is the vibrations frequecy shift due to the cabinets higer resonance therefore generating tones in the midrange causing midrange blurring |
|
|
| Svante |
| quote: | Originally posted by lopan
People seem to be talking about sub boxes as is they only resonate within the subs pass band (cabinet breathing). The arger issue is the vibrations frequecy shift due to the cabinets higer resonance therefore generating tones in the midrange causing midrange blurring |
This does simply not happen if the walls are reasonably linear. If the exciting signal does not contain the frequency of the resonance, there is no problem. |
|
|
| bwbass |
| FWIW, acoustic guitars, which function like very poorly made bass-reflex boxes, typically have a Helmholz air resonance around 20% lower than calculations would predict. This is due to the top and back flexing... the stiffer the top and back the higher the resonant frequency. I agree, though, that in most decently constructed sub boxes the frequency shift would likely be unmeasureable, but wall stiffness may have a slight effect nonetheless. |
|
|
| lopan |
I have an accelerometer somewhere around here so I will try to fine it and scope my subs whivh are 6.25 cubic feet divided in half made of 1.5 " MDF whith 36 internal braces. I know what I feel at the box is not of the signal frequency. Additionally two variovents and 13 lbs of wool and 2 30w-100s
Furthermore in a system in which multiple frequecies are introduced additional tones are generated by the sumation and difference of the initial tones and the generated tones
f1 + f2 +f3, f1+f2-f3, f1+f2, f1-f3, ... |
|
|
| paulspencer |
| quote: | | Even if your panels don't hit a resonance, if your total panel area is large compared with the driver effective cone area it won't take very much flexing of the panels to almost equal (and therefore undo) the air displacement of the driver. The panel area of mine is about 52 times that of the 12" cone. |
I'm very much in favour of making very rigid boxes but this sounds overstated.
eg. let's look at a 1 cu ft sealed sub with a 12" driver like the Stryke AV12.
VD is 2.3L.
The surface area of the box is 0.7m2.
For the box to move as much air as the driver requires the walls to move on average 1/14th of the p-p xmax of the driver at most, which is
1/14 x 46mm = 3.3mm
1.65mm ~ -6db
0.8mm ~ -12db
0.4mm ~ -18db
0.2mm ~ -24db
0.1mm ~ -30db
The walls would have to be very thin for you to be able to see them flexing this much.
In circlotron's box, the box has 3.7x more surface area, so it would need to be more rigid to achieve the same effect, however it seems hard to credit that the contributation from the box could be close to that of the driver.
The theoretical maximum output of the box should be 6db. This is output from the rear of the driver. If you achieve any more than this, then the rear of the box is acting as an acoustic lever. In the case of a sealed box, you could consider this in terms of transmission loss (TL) through the enclosure material. I would think that as long as you can achieve 10db TL without significant coloration, then this should be acceptable.
Most of us diyers like to go well beyond acceptable. We like to put our hands on the box and feel nothing when the speakers are playing. Many of us have a love of over-engineering. This is why I prefer to make boxes a minimum of 36mm thick ;) |
|
|
| inertial |
(sorry for my english)
Hi paulspencer,
right idea , the "double symmetrical closed box".
I have done one some years ago:) and the result is CATASTROPHIC:
Below 100 Hz the transmission loss is near to zero!
After, I "redirected" all my energy in the mechanics field.
Big work for next 50 years!!
best regards from Venezia |
|
|
| paulspencer |
| quote: | | Below 100 Hz the transmission loss is near to zero! |
When I have looked at the TL of different materials, it drops off significantly below 100 Hz. You make a good point here.
Has anyone seen data on the TL of 3/4" MDF below 100 Hz? |
|
|
| inertial |
Thank you, Paul .
I agree. |
|
|
| paulspencer |
I've been doing some hunting after inertial's comments. This turned up:
Figure 1. Transmission losses of typical single-leaf walls, A: 16 mm plywood, 10 kg/m², STC 21; B: 13 mm wallboard, 10 kg/m², STC 28; C: 1.3 mm steel, 10 kg/m², STC 30; D: 100 mm concrete, 235 kg/m², STC 52.
(http://irc.nrc-cnrc.gc.ca/cbd/cbd239e.html)
Notice how the 16mm plywood and the 1.3mm steel is about the same? Not what I would have expected. It has got me thinking about the possibility of using a sandwich laminated panel of sheet metal and 3mm MDF for future planned curved subwoofer walls ...
As you can see in the chart, the TL drops below 10dB and this is a bit or a worry! Concrete 100mm thick is approx 30dB.
As far as TL goes, according to one site "The best barriers are heavy, high mass, limp, highly damped materials with a high weight to stiffness ratio such as sheet lead or mineral loaded polymeric vinyl." (http://www.domesticsoundproofing.co.uk/tloss.htm)
"At low frequencies the stiffness of the material is the main controlling factor. Just above this point, various resonances cause major variation in sound transmission. "
This demonstrates why bracing is needed in subwoofers - stiffness control.
Here's a chart which is interesting:
Figure 2. Effect of air space on ideal double walls with 0.5 mm steel on each face, sound absorbing material in the cavity and no rigid mechanical connections between the faces. A has an airspace of 100 mm, a resonance dip at 135 Hz, and an STC of 29; B has an airspace of 5 mm, a resonance dip at 630 Hz, and an STC of 24. Curve C represents mass law predictions for a single 1 mm steel sheet and has an STC of 28.
This is effectively "constrained layer damping" and what is interesting is that it only works in the mid to high frequencies! It appears to have no benefit for bass! This is ignoring the benefit of CLD as vibration control. Vibration control from the driver and transmission loss are two different things. |
|
|
| inertial |
(sorry for my english)
Hi Paul,
Well done!
It is time to serious things! :)
These are normal diagrams , BUT I see "IDEAL", "IDEAL", "IDEAL".
I assicure all the interesting people, in the real behaviour all it is
MUCH WORSE!
Who is researching in this field ? Maybe High-end audio designer?
By |
|
|
| paulspencer |
| quote: | | These are normal diagrams , BUT I see "IDEAL", "IDEAL", "IDEAL". |
They appear as based on measured data to me. Perhaps they have been smoothed a lot, but you get an idea.
They don't really show bass, but if you extrapolate the curves, you get a fair idea. Better performance might be achieved, however, due to bracing, as this is data intended to be used for noise reduction in rooms, where the materials span further.
The attached file gives an idea of how transmission loss through a material is different through different parts of the audio bandwidth. In a subwoofer, the mains concerns would appear to be mass and stiffness. You can see how with a fullrange speaker, things are different. In the resonance controlled region, a speaker box made from solid materials with a high Q and low daming (hardwood rather than softer timber or MDF, or metal or plastic perhaps) would suffer in the lower midrange.
In the midrange, damping in the speaker box has more effect. I'd say the midrange is perhaps the area most likely to suffer from box coloration, but in this range it is easier also to have a more inert box.
It all becomes more complicated where you have to also consider both transmission loss and the vibration of the box via the mechanical coupling of the driver to the box. This is where I see value in attaching the driver to an inner box, then placing another box around it with a damping layer in between - constrained layer damping. If you built a push pull arrangement with the drivers on opposite sides of the box coupled to a constrained layer vibration damping design with matrix bracing on the inner box, then you would have a box that would perform very well. Use a sealed box with a pair of XBL2 drivers and you would have something very accurate. |
|
|
| paulspencer |
Here's another question to add to the discussion:
How does the alignment affect the construction of the box?
sealed, vented, TL, etc
A professional speaker builder once said to me that it's not necessary to make a vented subwoofer box as thick as you might with a sealed box. I'm not sure I agree. What are your thoughts? |
|
|
| inertial |
(sorry for my english)
Hi Paul,
" Mass law" presume much ideal semplifications.
Anyway, the last diagram show TL caused by aerial waves .
Structural transmission is much more efficient ( and complicated)
( how you said) .
For max stiffness :
a) shape
b) size
c) material
These determinate the numbers of vibration modes , etc,etc.
What is most stiffness shape? ( easy :) )
About thick for closed or bass-reflex: this is never enough!!
About two opposite wf .....
I think this: Imagine your hand between two hammers ( one from left
and one from right) Could be the result = zero? I think not. :)
regards |
|
|
| paulspencer |
| quote: |
I think this: Imagine your hand between two hammers ( one from left
and one from right) Could be the result = zero? I think not. |
Not a good illustration! You had will be crushed on both sides, but this is different!
In high school physics the concept of wave cancellation was demonstrated by having waves move towards each other from opposite ends. At the moment when the waves pass, they cancel each other out if their polarity is opposite. I believe this is what happens with mounting drivers opposite each other on a box - one on front, one on back. |
|
|
| inertial |
(sorry for my english)
Oh,incredible!
Imagine two identical earthquake one at north and one at south,
are the city in the middle intact?
What type of waves are you speaking ?
Put your 2 wf at the opposite side of a tube. It vibrate!
Structural transmission are not a my opinion but if you can show
the opposite you are welcome.
By |
|
|
| Ron E |
| quote: | Originally posted by paulspencer
In high school physics the concept of wave cancellation was demonstrated by having waves move towards each other from opposite ends. At the moment when the waves pass, they cancel each other out if their polarity is opposite. I believe this is what happens with mounting drivers opposite each other on a box - one on front, one on back. |
Suppose you have two woofers on opposite sides of the box Deliver a pulse to each, pushing the speakers "out" for an instant, this will push the baffle "in" for an instant on each side, This inward wave will pass around the box, and if there is symmetry and no damping, the waves will add when they meet. This is what "inertial" is talking about.
If you take a box and mount a driver on one side, another box and mount the woofers antipodally and put it on a track that constrains motion in the direction of the cones' axis, then measure the vibration of the cart, it will be less with a push-push system with antipodal woofers. THis sort of vibration is what is reduced in this manner. This is reaction vibration, not box vibration, per se...
Things are more complicated than that to judge overall vibration. Using two woofers means a bigger box with bigger panels which have a lower resonant frequency. Bigger panel area means more overall force on the panels for a given pressure change in the box and more panel vibration given the same contruction. The bigger you make the box, the more you have to brace it, regardless of how you mount the woofers. |
|
|
| inertial |
Thank you" Ron E" for your post.
Now I haven't time to reply, tomorrow I hope ......
( if the argument interest some people )
best regards |
|
|
| BassAwdyO |
| I mentioned this earlier. Push push only cancels "whole box vibration" Panel vibration is an entirely different thing. When the entire box vibrates due to inertia forces from the moving cone of the speaker I would imagine it acts as a dipole. The entire box moves slightly one way then the other following the signal of the speaker to some degree. The vibration can easily be coupled to the floor (in down or upward firing subs) or a side firing design can be used which will couple alot less. Ports should have a similar effect. If you have one port on the box the air mass moving in the port should have a slight inertia to it that pushes and pulls on the box. About ported boxes needing less panel thickness, I dont think thats true at all. In my experience a small hole such as a screwhole that was misplaced, will whistle more in a vented box at resonant frequency than a sealed box. This would lead me to believe a higher pressure difference between the inside and outside of the box and more panel flexing. I may be wrong here but my idea why this occurs is because the port resonance creates higher peaks and dips in the pressure wave at or near resonance. Anyone else have ideas on this? |
|
|
| paulspencer |
| If you consider the vent as a controlled tuned leak, then the pressure would be less overall. At least, intuition suggests a fully sealed box would have more pressure for the same cone excursion. However, the compound the confusion, vented boxes reduce cone motion near tuning. Perhaps someone can clear this up! |
|
|
| BassAwdyO |
| the tuning is where i think pressure increases in a vented box versus a sealed box. Thats why the excursion would decrease, think about it... If the pressure is higher in the box then when the cone moves in it is stiffened more, if the pressure is lower in the box then when the cone out it is stiffened more. Therefore the motion would be decreased. The port and air in the enclosure resonate and this is what causes the increase and decrease in pressure inside the box. I guess I still have no scientific proof to show for it. |
|
|
| lopan |
| quote: | | If you consider the vent as a controlled tuned leak, then the pressure would be less overall. At least, intuition suggests a fully sealed box would have more pressure for the same cone excursion. However, the compound the confusion, vented boxes reduce cone motion near tuning. Perhaps someone can clear this up! |
you are probably confusing resonace of the cone with resonance of the system (box tuning). At resonace the movement thru the port is at a maximum and the output is in phase. Therefore the box offers greater resistance to the cone movement. As the cone moves in the box pressure is greatly increasing due to port intake and out is the opposite.
Try this put a mass on a long rubber band and hold it with you your hand and mone your hand up and down till you find resonance not e that the mass movement (port volume) is at a max and the resistance to your hand movement is alo at a max (cone movement)
Or take a driver and attach a spring mass to it that matches the box and watch for a peak in the drive current.
for a spring mass f = 1/2pi*root(k/m) wher k is spring constant and m is mass and for a helmholtz k = p*a*a*v*v/V and m=pal where p is the air density, a is the port area, v is the velocity of sound in meters and V is the volume of your box in I believe liters and l is the length of the port witch must be adjusted for the open ends.
May be some one understands this better |
|
|
| m0tion |
| Posted a guide explaining how I built my new subwoofer (which uses the discussed "CLD" construction) on my website (sig). |
|
|
| Mjr7531 |
| I don't know if this was posted before, but I read an article somewhere where a guy worked with a braced box and an unbraced box and by bracing it, the group delay went down for the subwoofer. |
|
|
| steelyfan |
| quote: | Originally posted by Kittle
Stuffing a box always seemed counter productive to me.. we work so hard to get the internal volume correct (subtract port size, subtract speaker cone displacement, subtract bracing, etc...) - and then said volume gets reduced by some arbitrary amount of stuffing.
someone edumacate us please! |
TMK moderate stuffing does not reduce, but virtually enlarges the box volume. Explanation about it can be found in the book by mr Dickason.
KR KRIS |
|
|
| paulspencer |
IMHO us diyers are abbrev-obsessed and AFAIK this doesn't occur as much on other types of forums, though OTOH IIRC maybe diff on PC forums, BTW YMMV but SWMBO will probably agree on this ...
Are we talking about stuffing or lining? It's not a bad idea to put in a small amount of lining (say open cell foam) on the walls and this won't have any noticeable change to volume. Light stuffing with something like dacron is another matter. |
|
|
|
