• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Looking for insight into box/bracing design for 2.5-3 ft3 sealed subwoofer

HighFutility

Active Member
Joined
May 8, 2022
Messages
135
Likes
125
Hey, I was looking for insight into a new subwoofer build for myself. I have built five subwoofers. The earlier projects were not the greatest. Lack of tools, knowledge, and skill. I have come a long way and would like to tackle a bigger project for several reasons. All my subs have been 10-inch builds, some painted, some veneered.

I recently picked up a GRSSSW-4HE 15-inch subwoofer.
Link: https://www.parts-express.com/GRS-1...2NpS7UkJEJD5EoWESEw5jYIgXFIcAqbjflUEI7cEEe--q

I purchased this for several reasons, in no particular order: it was cheap and an introduction to 15-inch drivers. It was cheap enough that once my wife finds out about it, I wouldn't feel so bad when it came time to get rid of it. It's going to obviously be in the way and an eyesore. I just want to build it and see how it sounds and feels in my setup. The other part is im just itching to build something with my hands. I LOVE the process and learning new things along the way. This will also be my first build using a plate amplifier. All my other builds have been passive and powered by my crown xls 1502. Given the driver is listed with 300W RMS, I got a plate amp from Dayton that could drive it to its full extent.

Link to plate amp: https://www.parts-express.com/Dayto...-Subwoofer-Plate-Amplifier-300-806?quantity=1
I have questions regarding the 6db BOOST. More on that later.

I hope to use the forum members' expertise to make the best design possible.

For a sealed enclosure, Parts Express lists a 3 ft3 enclosure for a sealed F3 OF 36Hz. Booting up WinIsd I can see powering the full 300W it hits 12.5mm excursion Xmax is listed 12.3mm. If I build the box to 2.9 ft3 powering it full 300W I can stay below X max. Both 2.9 and 3 ft3 boxes listed below.
1728572502927.png


This is my current Dyaton 10-inch Ultimax 2 subwoofer build with 336W* (see next sentence) vs my upcoming GRS project build with 300W. @ 20hz it's 92 vs 100 spl, @ 30 hz 98 vs 107spl.
The crown xls is advertised as 525 into 4OHM Link: https://www.crownaudio.com/en/products/xls-1502; however, when Amir reviewed it link : https://www.audiosciencereview.com/...-and-measurements-of-crown-xls-1502-amp.6062/ . It showed 336W before clipping with a 4ohm load. When modelling system input power. Should I input 336W, or is the 525W manufacturer still valid? The Ultimax 10-inch driver can handle 600RMS. This is off-topic but valid in the modelling process.

1728572915145.png

Returning to the GRS BOX DESIGN and the plate amp with 6db boost @ 30hz. The amp is rated for 300W. Where does the power come from for the 6DB boost? Does the amp turn down the system input power by a factor of 4 to 75W in order to apply the 6db boost? If so, with an input power of 75W and a 6db boost @ 30hz (I have no idea of the q factor of the plate amp). We are back at square one. Image 300W vs 75W + 6db boost @ 30hz.

1728574018647.png

Kinda lame if thats the case.

So, is the best box size 2.9 ft3? I can drive the sub full tilt with the amp without hitting excursion. I can technically build the box smaller, but given the OD of the driver is 15.36 inchs, the smallest H x W I could get is 16.6 inches. I think I would need to leave at least an inch or more around. I plan to double baffle. While on the subject matter. Is there a rule of thumb for how much space to leave around the subwoofer driver? I see SVS subs look like they build the SB series really close leaving minimal space.

Bracing

I plan to use a double baffle. I came across this post link: https://www.stevemeadedesigns.com/board/topic/206664-implementing-bracing/ and thought of either using a donut brace or a window brace. Any recommendations for best practices would be welcome.
1728575290259.png
1728575265045.png


Here are few pics of my last build:
 

Attachments

  • hmRC_o8q.jpg
    hmRC_o8q.jpg
    103.5 KB · Views: 49
  • 5BGT8ZrF.jpg
    5BGT8ZrF.jpg
    104.8 KB · Views: 52
  • IKViUaBZ.jpg
    IKViUaBZ.jpg
    86.5 KB · Views: 48
The crown xls is advertised as 525 into 4OHM Link: https://www.crownaudio.com/en/products/xls-1502; however, when Amir reviewed it link : https://www.audiosciencereview.com/...-and-measurements-of-crown-xls-1502-amp.6062/ . It showed 336W before clipping with a 4ohm load.
Since we can't trust Crown I don't think we can trust any manufacturer... :( It all gets very complicated and "statistical" anyway because speakers are rated for peak program wattage. i.e. A 100W speaker is supposed to be safe with an amplifier hitting 100W on normal program material. But of course not all program material is equal. Dynamic music has a lower average level and it heats-up the voice coil less than highly-compressed music. Or if you push the amplifier into clipping, the average power goes-up (again more heating of the voice coil) even though the peaks are clipped.

Where does the power come from for the 6DB boost? Does the amp turn down the system input power by a factor of 4 to 75W in order to apply the 6db boost?
It's just boosting the gain. If you are near clipping without the boost you'll have to lower overall volume. And it depends on the program material. If there is no 30Hz sound, it won't have any effect and if there is weak sound at 30Hz it won't have much effect on the overall power. But the "theory" is to make-up for speaker roll-off so I'd expect less available power at the higher frequencies.

For a sealed enclosure, Parts Express lists a 3 ft3 enclosure for a sealed F3 OF 36Hz. Booting up WinIsd I can see powering the full 300W it hits 12.5mm excursion Xmax is listed 12.3mm. If I build the box to 2.9 ft3 powering it full 300W I can stay below X max. Both 2.9 and 3 ft3 boxes listed below.
I'd focus on frequency response. And I think a smaller sealed box just means you need more amplifier power to hit Xmax for the same SPL loudness.

Is there a rule of thumb for how much space to leave around the subwoofer driver?
With bass wavelengths it doesn't matter.

I don't know what's the best bracing. As long as it's stiff and the cabinet walls don't resonate you should be OK. Even if the cabinet walls vibrate a little, the woofer vibrates a LOT more and almost all of the sound comes from the woofer. It's probably more important (or equally important) to have some lining/stuffing inside to prevent resonances related to box dimensions. But with a subwoofer the box-size resonances are above it's operating range and generally won't be excited/activated anyway. (100Hz has a wavelength of about 10 feet.)
 
Internal bracing certainly helps but is limited. Be aware that the box can be 'braced' even further using 'External bracing'. The box could be built using two layers of MDF instead of just one using SOFT glue to sandwich them together. And/Or... pull out my trusty tile wet saw and then, using a soft urethane glue, apply 3/8" granite tiles to the exterior of the box. Granite tile can be had pretty inexpensively. Far stronger than just internal braces. Looks great, very stiff and solid. Still can be slid around the floor using appliance discs.

1728580173595.png
 
As already mentioned,
a smaller sealed box just means you need more amplifier power to hit Xmax for the same SPL loudness
at low frequencies. For example, GRS 15SSW-4HE in 100 l, 100 W(gray) and in 25 l, 800 :rolleyes: W (black), half-space:
1728587549264.png
1728587565310.png

1/2" Xmax at 70% BI (Klippel verified) seems too optimistic, for normal hifi use it can be divided by two.
Boost may not be needed in the room (I assume you will be using a sub in room). I would take into account room transfer function. Maybe real data will help https://www.audiosciencereview.com/...underdriven-pa-subwoofers.57216/#post-2095232 and below.
 
Last edited:
There seems to be a crusade for the very lowest achievable frequency. Lower must be better. My personal preference is that I find a sealed sub sounds better for music. I run the crossover low around 45hz and then push the volume up. I like the punchiness and texture it provides. To each their own.
 
Subs rarely needs bracing. Frequencies are too low to be a potential problem.
 
Subs rarely needs bracing. Frequencies are too low to be a potential problem.

Well, it's useful to break up panels and move resonance up in frequency?
 
Subwoofer frequencies do not cause internal resonances or problematic panel resonances.

I like overengineering. :) Our subs are -6dB at 300hz, and then in addition you get harmonics when you play loud, so theoretically you can excite a panel! :D
 
Threaded rod internally with inserts into the cabinet and the whole thing can be tensioned is a good method I used in some DIY speakers. Another was to cut out thick mdf, and stack it to form the cabinet. I've done that with both glue as well as adding end caps that used threaded rod to clamp the whole thing together top to bottom.

What I did was very much like this Magico speaker. I did it 30 years ago before they did. You can create complex braces and by layering change thickness and shape as you go. Doesn't have to be a rectangle.
1728595983754.png
 
The crown xls is advertised as 525 into 4OHM [...] however, when Amir reviewed it [...] It showed 336W before clipping with a 4ohm load.
Crown specs the power at 0.5% THD, while the 336W figure (or 388W in the next plot down) is approximately where the distortion just starts to rise sharply (onset of clipping). The power output according to Amir's measurements is 336W at 0.008% THD, 388W at 0.009% THD, and just short of 500W at the manufacturer-specified 0.5% THD.
 
My earlier concern with bracing was that the box is large, at least from my historical builds. Most of my subwoofers are around one ft3. This will be three times as big. I thought the panels might sag or be flexed due to the larger panel sizes. 300W = to 0.4 horsepower, I guess due to the lowest frequencies being 20 hz or 20 m/s the exerted forces are ~ 1.6 lbs? My math might be off.

Edit One: Math was off. Its about 3.2 lbs of force 20 times per second. Is that not enough force to flex a panel? That is 16.5 wide by 20 inches long?

Edit two: I'm not worried about resonances; it's just that my box frequency response is changing, potentially coming undone at the seams. I know some subs have 1000W 5000W+, and for every 750W= 1 horsepower, I thought bracing was more for preventing flex than resonances.
 
Last edited:
This is a cab for a 3way that I designed and built. FB=38hz. 90Litres. 300hz cross for planned system but could be a sub box. The ports have sealed covers for use as a midbass cab in a 4 way

1/4" BB side walls with 1/4" thin Spalted thin boards. My goal was to make it as light as possible. -stiff and light-

The drivers are mass loaded to the box by coupling them to the bracing for a level of dissipation through entire box materials. Our forefathers(maybe 1950's) would hang baffles from the ceiling with a gasket seal from baffle to box side walls for decoupling which is more effective - I have no plans for that method!

Box ratio has 3 different dimensions and somewhat golden!
Side braces divided into 3 separate dimensions.

Through the rear holes the ports can be adjusted from 5 to 8 inches to produce the alignment range as shown, dependent on room gain and or personal choice.. Range shown in attachment
 

Attachments

  • Port Adjustment.PNG
    Port Adjustment.PNG
    49.9 KB · Views: 39
  • thumbnail_20221016_114253.jpg
    thumbnail_20221016_114253.jpg
    124.2 KB · Views: 42
  • thumbnail_20221016_114223.jpg
    thumbnail_20221016_114223.jpg
    92.3 KB · Views: 43
  • thumbnail_20221002_172806.jpg
    thumbnail_20221002_172806.jpg
    188.8 KB · Views: 42
  • Brace round holes.PNG
    Brace round holes.PNG
    918.2 KB · Views: 37
Last edited:
Here's the sub that I recently built... I agree with Sigberg on the overengineering comment. I don't want any part of my box moving around, even theoretically, and since it falls into the hobby realm, the extra time and effort required to do it is part of the fun. :)

IMG_4165.jpg


Doing it with braces that run from front-to-back, all the way from the lip of the sub, means that in the end, the sub is pushing on the whole box, rather than just the front baffle. For this front baffle to flex, the sub is going to have to be pushing hard enough to physically compress solid wood.

IMG_4166.jpg


Here's what the bracing looked like once it was glued. This is just before I slathered the entire interior in automotive anti-resonance rubberish paint.

IMG_4183.jpg


No bearing on the bracing, but since I'm posting pics, here's what it looked like from the outside after glue up. Took 10x longer to make than if I had gone the square box route, but what I lost in time I gained in multiples when it came to WAF. I have not heard a single grumbling comment from the powers-that-be on these (or the speakers that go on top, which are generally similar), even though they objectively take up more space than my old monkey coffins. ;)
 
Here's the sub that I recently built... I agree with Sigberg on the overengineering comment. I don't want any part of my box moving around, even theoretically, and since it falls into the hobby realm, the extra time and effort required to do it is part of the fun. :)

View attachment 398063

Doing it with braces that run from front-to-back, all the way from the lip of the sub, means that in the end, the sub is pushing on the whole box, rather than just the front baffle. For this front baffle to flex, the sub is going to have to be pushing hard enough to physically compress solid wood.

View attachment 398064

Here's what the bracing looked like once it was glued. This is just before I slathered the entire interior in automotive anti-resonance rubberish paint.

View attachment 398065

No bearing on the bracing, but since I'm posting pics, here's what it looked like from the outside after glue up. Took 10x longer to make than if I had gone the square box route, but what I lost in time I gained in multiples when it came to WAF. I have not heard a single grumbling comment from the powers-that-be on these (or the speakers that go on top, which are generally similar), even though they objectively take up more space than my old monkey coffins. ;)
Your build is beautiful but exceeds my skill and tool's capabilities.
 
As already mentioned,

at low frequencies. For example, GRS 15SSW-4HE in 100 l, 100 W(gray) and in 25 l, 800 :rolleyes: W (black), half-space:
View attachment 397881View attachment 397882
1/2" Xmax at 70% BI (Klippel verified) seems too optimistic, for normal hifi use it can be divided by two.
Boost may not be needed in the room (I assume you will be using a sub in room). I would take into account room transfer function. Maybe real data will help https://www.audiosciencereview.com/...underdriven-pa-subwoofers.57216/#post-2095232 and below.
What do you mean by /2" Xmax at 70% BI (Klippel verified) seems too optimistic, for normal hifi use it can be divided by two.


This thread is about a subwoofer im looking to build. How do I take into account room transfer function>? Is that the same thing as room gain?

I am new and not familiar with everything.
 
Threaded rod internally with inserts into the cabinet and the whole thing can be tensioned is a good method I used in some DIY speakers. Another was to cut out thick mdf, and stack it to form the cabinet. I've done that with both glue as well as adding end caps that used threaded rod to clamp the whole thing together top to bottom.

What I did was very much like this Magico speaker. I did it 30 years ago before they did. You can create complex braces and by layering change thickness and shape as you go. Doesn't have to be a rectangle.
View attachment 397924
If I had access to a CNC machine and unlimited time/budget I would be there. Unfortunately, I have some hand tools and the ability to stay within my means.
 
Crown specs the power at 0.5% THD, while the 336W figure (or 388W in the next plot down) is approximately where the distortion just starts to rise sharply (onset of clipping). The power output according to Amir's measurements is 336W at 0.008% THD, 388W at 0.009% THD, and just short of 500W at the manufacturer-specified 0.5% THD.
How did you get the figures down to a single digit from the graph? Is there an Excel file with more detail?
 
Back
Top Bottom