I’ve been building subwoofers for years now, and I’ve always used car audio drivers. However, I recently started building speakers using pro-audio drivers, specifically mid-woofers and horn loaded compression drivers. I love how dynamic and clean-sounding these drivers are, so it was only natural that I decided to give pro-audio subwoofers a try.
I did some research on pro-audio subwoofer drivers with relatively low Fs (resonant frequency) and high Xmax (maximum linear cone excursion), and the Lavoce SAF184.03 stood out as a promising candidate. It features a beefy 4" voice coil and 1500W power handling. Some might scoff at the Lavoce’s 30 Hz Fs and 13 mm Xmax, but keep in mind that I plan to build multiple subwoofers in tiny sealed enclosures – at least by 18" subwoofer standards, so deep bass output likely won’t be an issue, and I’ll still be able to keep enclosure sizes in check.
I eventually settled on an enclosure design featuring generous 4" radius rounded corners even though I knew it would be somewhat challenging to make. Using 15 mm kerf-bent Baltic birch plywood not only keeps the enclosure weight in check but also is stronger than MDF, not to mention the wood grain looks very nice when stained a light color. The braces and baffle were made from 18 mm MDF.
I used a biscuit joiner to cut slots in the plywood. The two MDF braces are attached to the plywood using biscuits, essentially serving as the form around which the plywood was bent.
The braces also feature circular openings to support the heavy subwoofer motor, which probably weighs close to 15 kg (~33 lbs) by itself.
The front baffle and rear panel were painted black, and the birch plywood was stained white with a two-component oil. I lined the walls with polyfill sheets.
I placed four pillows filled with polyfill inside to improve deep bass response. In total, I used about 1 lb of polyfill per cubic foot. The gross internal volume, before accounting for driver displacement, is approximately 3 cubic feet (~85 liters), which is about 25% smaller than the typical minimumenclosure size recommended for a standard 18" car/home audio subwoofer.
With the addition of polyfill, the subwoofer’s resonant frequency decreased from about 61 Hz to 56 Hz. The impedance curve reveals that inductance is well-controlled, allowing this subwoofer to be easily crossed over as high as 500 Hz. The nominal impedance is 8 ohms.
The near-field frequency response indicates a roll off that’s less steep than the expected 12 dB/octave for a sealed enclosure, thanks to the driver’s strong motor. When placed along a wall in a typical sized room, the subwoofer should deliver flat frequency response down into the mid-twenties, especially after reducing the ample mid-bass output using EQ or an AVR’s room correction capabilities.
Finally, a complete build video is available below. Please let me know what you think.
I did some research on pro-audio subwoofer drivers with relatively low Fs (resonant frequency) and high Xmax (maximum linear cone excursion), and the Lavoce SAF184.03 stood out as a promising candidate. It features a beefy 4" voice coil and 1500W power handling. Some might scoff at the Lavoce’s 30 Hz Fs and 13 mm Xmax, but keep in mind that I plan to build multiple subwoofers in tiny sealed enclosures – at least by 18" subwoofer standards, so deep bass output likely won’t be an issue, and I’ll still be able to keep enclosure sizes in check.
I eventually settled on an enclosure design featuring generous 4" radius rounded corners even though I knew it would be somewhat challenging to make. Using 15 mm kerf-bent Baltic birch plywood not only keeps the enclosure weight in check but also is stronger than MDF, not to mention the wood grain looks very nice when stained a light color. The braces and baffle were made from 18 mm MDF.
I used a biscuit joiner to cut slots in the plywood. The two MDF braces are attached to the plywood using biscuits, essentially serving as the form around which the plywood was bent.
The braces also feature circular openings to support the heavy subwoofer motor, which probably weighs close to 15 kg (~33 lbs) by itself.
The front baffle and rear panel were painted black, and the birch plywood was stained white with a two-component oil. I lined the walls with polyfill sheets.
I placed four pillows filled with polyfill inside to improve deep bass response. In total, I used about 1 lb of polyfill per cubic foot. The gross internal volume, before accounting for driver displacement, is approximately 3 cubic feet (~85 liters), which is about 25% smaller than the typical minimumenclosure size recommended for a standard 18" car/home audio subwoofer.
With the addition of polyfill, the subwoofer’s resonant frequency decreased from about 61 Hz to 56 Hz. The impedance curve reveals that inductance is well-controlled, allowing this subwoofer to be easily crossed over as high as 500 Hz. The nominal impedance is 8 ohms.
The near-field frequency response indicates a roll off that’s less steep than the expected 12 dB/octave for a sealed enclosure, thanks to the driver’s strong motor. When placed along a wall in a typical sized room, the subwoofer should deliver flat frequency response down into the mid-twenties, especially after reducing the ample mid-bass output using EQ or an AVR’s room correction capabilities.
Finally, a complete build video is available below. Please let me know what you think.
That is very nice work. And also good production. These supercompact close enclosure make a lot of sense with these drivers. Great value for the displacement you get. I also often look into car audio drivers, but after covid, less and less great deals are to be scored, so PA is winning momentatily at least.
Wow you did a very nice job with that! How deep & spaced were your cuts? What was the thickness of the blade?
Thanks! The blade thickness is just 1.5 mm, and the spacing between cuts is about 9 mm. I cut through all but the last two plies. I used the kerf spacing calculator at https://www.blocklayer.com/kerf-spacing to determine the necessary spacing.
That's an interesting driver. Your cabinet looks great! The internal bracing looks very nicely executed too. The video is very interesting to watch.
I've gone and created a model of this subwoofer using VituixCAD. I used the driver's Thiele–Small parameters as found in the relevant datasheet.
By modelling the driver in sealed enclosure with an effective volume Vb = 114 litres and quite heavy damping of QA=5.7 and QL=5.7, I got the impedance curve shown below. It's a reasonable sort of match to your measured impedance curve.
The simulated SPL response for 35W re 8Ω of input power is shown below. It's been adjusted to match the SPL of your measurements at 20Hz. All in all, I think that the simulation agrees quite well with your response curve.
Thank you for sharing all of your measurements. 🙂If we add the typical 80Hz 4th-order Linkwitz-Riley low-pass filter we get the following response curve. This of course isn't −6dB at 80Hz.
Adjusting the low-pass filter frequency downwards to 57Hz produces the following result. Here we see that the subwoofer is −3dB at 30.4Hz, and the low-frequency roll-off rate is quite linear at 10dB/octave.
Now lets increase the power (i.e., open the throttle). With 500W re 8Ω, we just start to hit Xmax at 20Hz, and the peak output from the subwoofer is a decent 114dB at about 48Hz, and at 20Hz it's around 105dB.
Let's apply some +6dB of EQ at 20Hz with Q=1.0. The results are shown below. Of course, this is for an input power level of 120W re 8Ω, with a peak power requirement of 400W around 20Hz (no surprise there). f6 has been lowered from 24.1Hz to 19.2Hz, a drop of almost 5Hz. Of course, our maximumSPL is now about 110dB versus the earlier 114dB. I think that's a good trade-off. Note that the 4th-order Linkwitz-Riley low-pass filter was set to 61Hz for these simulations.
Thank you for taking the time to run the simulations and share your results. As you found, a mere 500W is enough to reach Xmax at 20 Hz without using any EQ. However, Lavoce specifies an additional 10 mm of excursion beyond Xmax, so there's a comfortable safety margin, albeit with increased distortion.
On the subject of the impedance curves, I was a bit surprised that I needed to use QA=5.7 and QL=5.7 in order to add a lot of damping to the closed-box system so that the simulated impedance curve matched up with the measured results. If I didn't, the impedance peak would have been much higher. The free-air impedance simulation matched up quite well with the published curve in the Lavoce SAF184.03 datasheet, so I don't know what's happening with the closed-box enclosure.
The impedance peak was indeed much higher before I added about 1 lb per cubic foot of polyfill. Here's a measurement.
Really nice, clean build! I like how you did the painted baffle on natural wood cabinet. How does it sound?
As an aside, I have been attempting something similar, combining a hard oil finish baffle with a painted cab on my current project, but still struggling a bit with the best way to achieve this. When I masked off the paint, the oil finish found its way past the tape and discoloured the paint. So instead I tried finishing the pieces separately and gluing together at the end. This resulted in clamp marks in the oil finish, despite trying to spread out and use the minimum pressure.
As an aside, I have been attempting something similar, combining a hard oil finish baffle with a painted cab on my current project, but still struggling a bit with the best way to achieve this. When I masked off the paint, the oil finish found its way past the tape and discoloured the paint. So instead I tried finishing the pieces separately and gluing together at the end. This resulted in clamp marks in the oil finish, despite trying to spread out and use the minimum pressure.
Thanks! I think it sounds quite clean, and the mid-bass punch is definitely there.
I used a two component oil finish (oil + hardener). It dries pretty fast. I think that prevents the oil from interacting with the paint.
I normally use a 100 Hz crossover as my DIY mains also use pro-audio drivers that start to roll off above that point. I recommend going as high as possible to take advantage of the subwoofer’s effortless mid-bass output, and using EQ or room correction software like Dirac to flatten the response.
Very nice work - it makes a refreshing change to see something different to a plain old cuboid. Most people have never heard the effortlessly clean bass of which a large PA driver in a sealed enclosure is capable.
Reproducing bass accurately requires a large radiating area as implemented here. Simply focusing on displacement volume is not the answer - a small high Xmax driver flapping away at it's limit in a resonant ported box is not the same as a large Sd driver just 'ticking over', no matter what the response measurements show.
Acoustic instruments such as a double bass or gran cassa have large radiating areas for a reason! There is a thing known as 'acoustic impedance matching' which describes how well the source (speaker) transfers energy into the air (load). The larger the radiating area, the more favourable the matching of source to load. The concept is very well understood with HF horns but the poor old subwoofer never seems to receive any consideration in this respect...
Reproducing bass accurately requires a large radiating area as implemented here. Simply focusing on displacement volume is not the answer - a small high Xmax driver flapping away at it's limit in a resonant ported box is not the same as a large Sd driver just 'ticking over', no matter what the response measurements show.
Acoustic instruments such as a double bass or gran cassa have large radiating areas for a reason! There is a thing known as 'acoustic impedance matching' which describes how well the source (speaker) transfers energy into the air (load). The larger the radiating area, the more favourable the matching of source to load. The concept is very well understood with HF horns but the poor old subwoofer never seems to receive any consideration in this respect...
Does the oil finish harden fully and allow the paint to be applied last to avoid the bleeding problem?
I actually applied the paint first and then the oil finish. It hardens pretty quickly, and I haven’t seen any issues so far. By the way, the paint is water-based.