Building a large subwoofer, trying to make it good but inexpensive to construct. My initial idea was make it double walled and sand fill between the walls. After reading up on the subject it seems I could get similar benefit from stiffness + any constrained layer, it doesn't have to be sand and it doesn't have to be thick/weigh as much. Also realized that I could save significant $ on the project if I use 7/16 OSB and more bracing vs 3/4 MDF or plywood. Finally, experiments by a forum member in another thread found that only the baffle and back wall contribute to sound quality; he was unable to find any measurable effect from varying the side walls.
Given all of the above, I'm now at the following plan: 7/16th OSB, double wall for baffle and back wall only. Side walls braced but not double wall. Using 2" x 3" studs for the spacing and bracing. The back wall would be constructed like a hollow core door. I plan to put layers of roofing felt between the stud core and each panel of the double wall. Still brace it on the inside too.
The only thing is that the simplest way to run the 2x3 "ribcage" that connects the front, back and sides and mounts the bracing inside the internal volume (i.e. now talking about the inside of the speaker not the hollow core) would have the ribs 15" apart. Is this sufficient?
This is for a big sealed bass enclosure with four 15" subwoofers, going to be about 5ftx3ftx2.5ft.
(Using a double wall hollow core baffle will allow me to mount 15"x4=60" of drivers in just 60" by having half facing out and half facing in, because the baffle will be so thick. I'm planning to cut reinforcing ring out of 3/4" MDF and the drivers will mount to those and those to the MDF.)
What do you guys think?
Given all of the above, I'm now at the following plan: 7/16th OSB, double wall for baffle and back wall only. Side walls braced but not double wall. Using 2" x 3" studs for the spacing and bracing. The back wall would be constructed like a hollow core door. I plan to put layers of roofing felt between the stud core and each panel of the double wall. Still brace it on the inside too.
The only thing is that the simplest way to run the 2x3 "ribcage" that connects the front, back and sides and mounts the bracing inside the internal volume (i.e. now talking about the inside of the speaker not the hollow core) would have the ribs 15" apart. Is this sufficient?
This is for a big sealed bass enclosure with four 15" subwoofers, going to be about 5ftx3ftx2.5ft.
(Using a double wall hollow core baffle will allow me to mount 15"x4=60" of drivers in just 60" by having half facing out and half facing in, because the baffle will be so thick. I'm planning to cut reinforcing ring out of 3/4" MDF and the drivers will mount to those and those to the MDF.)
What do you guys think?
Double wall vs more bracing: suppose I run the core studs long ways and the inner braces short ways. Even if the double wall is no stronger than putting all the bracing inside on a single wall nox (e.g. in a '#' arrangement) it still requires fewer cuts because in the # either the vertical or the horizontal braces have to be cut to allow the others past (unless you notch them both).
15" spacing between braces is too far apart. Aim for <6". Use cross bracing between opposing sides, combined with smaller corner braces connecting adjacent sides. If you have long crossbrace members, let them touch in the center to be glued together to prevent them vibrating. Invent a pattern of cross brace and diagonal corner braces like in a bridge support structure that fits into your box, easy to install, and puts attachment footprints so that there is no unbraced span bigger than 6". Forget about ribs along the walls, they do little to stop vibration. You need to make the walls work against each other. Pressure variation affects all panels the same. They all want to move in or out together. Connecting them with bracing prevents this movement. Panel ribs just bend along with the panel. Ribs are useful in the corners for gluing and screwing the outer plywood layer together, but if you use a very strong adhesive like Loctite PL urethane construction adhesive with good clamping then you don't need ribs or screws.
One crossbrace between adjacent drivers is adequate since their metal frames are extremely stiff and provide a lot of bracing to stop baffle flex, but the momentum of the heavy cone can cause baffle to flex if no cross bracing front to back.
Hollow core door will vibrate itself to dust under the vibration from 4 big woofers. If the internal door glue fails it will buzz. The thin plywood skins will buzz when their glue gives up. Use high quality (lots of plies) 3/4 plywood like marine plywood or genuine baltic birch (not the chinese version) on the outside layer. I don't know how well OSB will endure vibration, but it seems like it is not engineered for extended violent vibration, but more for price, impact stress and water resistance. I defer to your expertise on OSB! I know that genuine baltic birch ply from Russia or Finland is really, really strong and a joy to work with. Domestic marine ply also very good. I have built a lot with Columbia Forest Products Purebond plywood from HomeDepot in USA. It has some minor voids occasionally, but otherwise very good quality for the money $50 4x8 sheet with nice veneer.
Panel resonance isn't as important on true subwoofer <80Hz, because the resonant frequency band of wood is far above the sub signal band. So it will flex with the internal pressure, but it won't resonate. Correct bracing raises the resonant frequency even more.
Rather than all 4 drivers on the front baffle, why not install 2 each on opposing sides, so they oppose each other's vibration? It could be shorter, but probably deeper if 2 subs still face forward, or let them face sideways, whatever it doesn't matter acoustically. If it is a true subwoofer, not playing above 80Hz, then the sound radiation of each driver is spherical, there is no directionality, so it doesn't matter where they go on the box for sound quality purposes. But putting them opposite sides eliminates a lot of the vibration that you would otherwise have to brace against. Connect the opposite drivers together with inside threaded aluminum hex bar acting as nuts for the driver mounting bolts. Then use cross and corner braces on the remaining 4 sides of the box, corner braces connect the other 4 sides to the driver baffles.
What kind of drivers and amp have you in store for this beast?
One crossbrace between adjacent drivers is adequate since their metal frames are extremely stiff and provide a lot of bracing to stop baffle flex, but the momentum of the heavy cone can cause baffle to flex if no cross bracing front to back.
Hollow core door will vibrate itself to dust under the vibration from 4 big woofers. If the internal door glue fails it will buzz. The thin plywood skins will buzz when their glue gives up. Use high quality (lots of plies) 3/4 plywood like marine plywood or genuine baltic birch (not the chinese version) on the outside layer. I don't know how well OSB will endure vibration, but it seems like it is not engineered for extended violent vibration, but more for price, impact stress and water resistance. I defer to your expertise on OSB! I know that genuine baltic birch ply from Russia or Finland is really, really strong and a joy to work with. Domestic marine ply also very good. I have built a lot with Columbia Forest Products Purebond plywood from HomeDepot in USA. It has some minor voids occasionally, but otherwise very good quality for the money $50 4x8 sheet with nice veneer.
Panel resonance isn't as important on true subwoofer <80Hz, because the resonant frequency band of wood is far above the sub signal band. So it will flex with the internal pressure, but it won't resonate. Correct bracing raises the resonant frequency even more.
Rather than all 4 drivers on the front baffle, why not install 2 each on opposing sides, so they oppose each other's vibration? It could be shorter, but probably deeper if 2 subs still face forward, or let them face sideways, whatever it doesn't matter acoustically. If it is a true subwoofer, not playing above 80Hz, then the sound radiation of each driver is spherical, there is no directionality, so it doesn't matter where they go on the box for sound quality purposes. But putting them opposite sides eliminates a lot of the vibration that you would otherwise have to brace against. Connect the opposite drivers together with inside threaded aluminum hex bar acting as nuts for the driver mounting bolts. Then use cross and corner braces on the remaining 4 sides of the box, corner braces connect the other 4 sides to the driver baffles.
What kind of drivers and amp have you in store for this beast?
Richidoo I like your suggestions! It's using the 15" Titan buy out drivers. I'm actually building two enormous 3-way speakers, each with 4 drivers / 8 for the set. I posted in the subwoofer forum, however, because I'm designing it from the lowest frequency up & the enclosure is of course dominated by the subwoofer volume. They're going to go in a very large room for salsa parties and I'm designing it on the principle of use inexpensive drivers but overbuild it so the drivers are working far below their theoretical max power. So putting the subwoofer drivers as side firing would make it easier to design the front baffor for mid/high. I might even do 4 way if it makes sense. I'm planning to use active crossover and separate amps for each driver; since I'm not using max power rating I can start out with inexpensive class D boards until I finish designing and building a discrete class AB amp.
I understand that you wanted to build sealed but if you went ported, ..you could consider sonotube subwoofers. The cylindrical shape is optimal for distributing the ppressure radially from the cylinder's axis.
Not sure if this is the best idea for a sealed sub though
OK I think that I will put them in opposing pairs. A couple questions: how do I model side firing opposing pairs in a speaker design program e.g. "Basta!" ? If I put the side firing speakers as close as possible to the front, will that allow me to cross over at a higher frequency to midwoofers in the front baffle? Or, what do you think of a design where the subwoofers are inset into the box and face to the side but additional walls guide the sound out slots in the front?
I originally intended to do 4 boxes but the owner of the space wanted 2; they can be big but not dramatic so something like a giant sonotube is probably out.
I originally intended to do 4 boxes but the owner of the space wanted 2; they can be big but not dramatic so something like a giant sonotube is probably out.
Hi denito,
Take a look here: http://www.diyaudio.com/forums/subw...nterested-ppsl-enclosures-55.html#post3993346 Post #543. Norman bates has a picture of a sealed PPSL subwoofer using 4ea. 12" drivers. Something similar could be done w/ 15" drivers. The PPSL method can be used to reduce the box front. what is the crossover frequency you are shooting for?
Regards,
Take a look here: http://www.diyaudio.com/forums/subw...nterested-ppsl-enclosures-55.html#post3993346 Post #543. Norman bates has a picture of a sealed PPSL subwoofer using 4ea. 12" drivers. Something similar could be done w/ 15" drivers. The PPSL method can be used to reduce the box front. what is the crossover frequency you are shooting for?
Regards,
Yes this looks similar to one of the concept drawings I made yesterday! The "donut".
Thanks for the CAD sketch and the tip about using half Sd.
What I was thinking of doing is similar overall dimensions as your drawing, same way of having the 4 drivers face, but move the drivers to the center and change the slot opening to the center of the top and the bottom plate. That will allow me to use the full height of the center of the front face for a line array of mid drivers. So, it will be a 5ft tall rectangular donut that is a down firing + up firing subwoofer.
I think this should be both incredibly strong and vibration-free because the forces from the inward firing drivers will be closely walled in on four sides in the donut hole, and the sound pressure coming out will be balanced top and bottom. No outer wall will be more than 9 inches from an inner wall so the bracing will be rock solid. Think of a balloon animal folded into a tight square, and how much more solid it's walls are than a round balloon of the same volume.
I can construct the skeleton/bracing as a "tire stack" of 8 identical square donuts, made from OSB, spaced every 7.5 vertical inches. Since the shape of the speaker will be encoded in the shape of the layers, it would take as much force as would bust a cofferdam to flex the walls at all.
Last edited:
Hi denito,
Post #14: "...it will be a 5ft tall rectangular donut that is a down firing + up firing subwoofer."
Will that result in a 5ft long center duct from top to bottom? That would probably cause resonance problems, and would loose a lot of internal volume, and with your choosen driver you need all the volume you can get. It might be better to use two relatively small plenums on top and bottom?
Would be great to see a sketch of your design idea.
Regards,
Post #14: "...it will be a 5ft tall rectangular donut that is a down firing + up firing subwoofer."
Will that result in a 5ft long center duct from top to bottom? That would probably cause resonance problems, and would loose a lot of internal volume, and with your choosen driver you need all the volume you can get. It might be better to use two relatively small plenums on top and bottom?
Would be great to see a sketch of your design idea.
Regards,
Yes, exactly! My next question was going to be, should the duct be straight-through or should each be a separate plenum. I hadn't realized a straight through might cause resonance.
Is there any problem with making the plenums have 90 degree corner walls instead of a segmented semicircle? The reason why is, if you imagine matrix bracing, such a matrix divides the enclosure into 3x3x3 boxes. (Like a nonsquare Rubix cube). The plenums would each occupy one sub-box bounded by the bracing matrix.
OK I found a voxel modeling program to make a sketch of it. I didn't use enough voxels so I couldn't show much detail.
Speaker enclosure shown without outside walls to show inside. Tan is OSB braces, orange is MDF plenum. White (between the orange) shows OSB panel closing off plenum hole. Light blue is 2x3 studs to join outside corners. Dark blue is integrated separate enclosure for front mid-woofers.
Speaker enclosure shown without outside walls to show inside. Tan is OSB braces, orange is MDF plenum. White (between the orange) shows OSB panel closing off plenum hole. Light blue is 2x3 studs to join outside corners. Dark blue is integrated separate enclosure for front mid-woofers.
Hi denito,
Post #16: "...Is there any problem with making the plenums have 90 degree corner walls instead of a segmented semicircle?"
The simulations of other PPSL enclosures show a response peak above the passband. That peak should not be a problem in a subwoofer, but that depends on how high you want to run it; also, according to djk (who came up w/ the PPSL design - http://www.diyaudio.com/forums/subwoofers/177905-thread-those-interested-ppsl-enclosures.html ) the simulations do not reflect the effect of the drivers taking up volume in the plenum, so any peak shown in the simulation will be much smaller. The short answer is: yes, square corners should also work, but the arrangement I show in my sketch in Post #15 might have less of an upper response peak. You also need to make the plenum big enough to get the speakers and your hands into it.
Your sketch in Post #18 shows, that a lot of the internal volume of your big box is taken up by wood and speakers. You need to take that into consideration when simulating the enclosure. If the inside gets to be too small you'll have a lot of woofers, but no woof. I'm assuming that you're going to use an electronic crossover to match the different speakers, so some equalization should be easy to design into that?
I just posted a number of links in another thread, they may come in handy:
http://www.diyaudio.com/forums/subwoofers/303524-what-does-eq-ing-subwoofer-entail.html Post #3.
Regards,
Post #16: "...Is there any problem with making the plenums have 90 degree corner walls instead of a segmented semicircle?"
The simulations of other PPSL enclosures show a response peak above the passband. That peak should not be a problem in a subwoofer, but that depends on how high you want to run it; also, according to djk (who came up w/ the PPSL design - http://www.diyaudio.com/forums/subwoofers/177905-thread-those-interested-ppsl-enclosures.html ) the simulations do not reflect the effect of the drivers taking up volume in the plenum, so any peak shown in the simulation will be much smaller. The short answer is: yes, square corners should also work, but the arrangement I show in my sketch in Post #15 might have less of an upper response peak. You also need to make the plenum big enough to get the speakers and your hands into it.
Your sketch in Post #18 shows, that a lot of the internal volume of your big box is taken up by wood and speakers. You need to take that into consideration when simulating the enclosure. If the inside gets to be too small you'll have a lot of woofers, but no woof. I'm assuming that you're going to use an electronic crossover to match the different speakers, so some equalization should be easy to design into that?
I just posted a number of links in another thread, they may come in handy:
http://www.diyaudio.com/forums/subwoofers/303524-what-does-eq-ing-subwoofer-entail.html Post #3.
Regards,
Okay! I've got a "real" model of it in Sketchup now! This just shows the bracing panels; there will also be battons/stick-frame where panels meet at 90 degree angles.
If you think of the internal structures as a "+" shape, the chambers formed by the front leg of the two "+" structures are for mid range speakers. The rear legs of the "+"s are the same dimensions, but with open backs - I'm re-using the same inner box shapes for bracing as I use for sub-chambers.
I'm not sure whether the un-braced sections of the outer walls are too wide, and if so which inner panels/battons I want to extend...
If you think of the internal structures as a "+" shape, the chambers formed by the front leg of the two "+" structures are for mid range speakers. The rear legs of the "+"s are the same dimensions, but with open backs - I'm re-using the same inner box shapes for bracing as I use for sub-chambers.
I'm not sure whether the un-braced sections of the outer walls are too wide, and if so which inner panels/battons I want to extend...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.