• 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!

Low distortion drivers for sub-bass in shallow sealed cabinets

And you think I'm crazy!

Not really, I just questioned if using shallow mount drivers was the best approach to the initial low frequency dB requirements. :)
 
The RedCat woofer is either a Peerless XLS clone, or a OEM Tymphany made driver.
 
No shallow mount driver is perfect, typically struggling with assymetric BL etc, but the SB acoustics one is a decent driver as long as you use enough of them. It's popular in shallow bass towers here in Norway.

Here is a video featuring yours truly assembling and tuning a bass tower with those drivers (from around 1:05). :)

Really nice finish on those cabs. But you know, SB makes those drivers in black, too. :)

Also curious, what's the cabinet volume per driver? One of the neat things about these drivers is that they play pretty well in tiny cabinet volumes. I've used them in as little as ~6L per driver!
 
Really nice finish on those cabs. But you know, SB makes those drivers in black, too. :)

Also curious, what's the cabinet volume per driver? One of the neat things about these drivers is that they play pretty well in tiny cabinet volumes. I've used them in as little as ~6L per driver!

I am aware that you can get them in black. I neither built or ordered the subwoofers, I just tuned them - apparently the customer wanted silver. :)

I don't remember the volume off the top of my head, but I'm guessing around 10-15 liters or so.
 
@jhaider And note that his main speakers are black with silver drivers too, so they match the look of those quite well.
 
I did some more analysis using both manufacturer specs and published reviews to derive 20Hz efficiency for all the drivers mentioned (including the Purifi for a non-shallow reference). It turns out that the very low Fs and fancy magnets of the SB make it by far and away the strongest performer, with 78dBSPL/2.83V/1m, while the purifi is just 67dB. System cost is therefore the lowest of any option due to the lower number of drivers required, even after allowing for distortion (probably) growing relatively fast. The Purifi grows to 8% distortion @ 72dB output (per hificompass.com), making it a rather bad choice for a sealed sub, whereas the SB is 5% @ 88.5dB (per AudioXpress), making it an excellent sub driver all round, not just for shallow designs.

Thanks everyone for all the insights! I will post my spreadsheet once I get back from holiday next week.
 
Here's the table, sorry can't seem to attach the actual table data. If any of the steps seem wrong do say. Since the numbers are based on using the drivers at the measured voltage, we can assume power handling is not an issue.

index.php
 

Attachments

  • 1727162695703.png
    1727162695703.png
    122.4 KB · Views: 235
  • 1727162911704.png
    1727162911704.png
    130.1 KB · Views: 28
@mcdn I haven't spent much time to understand what you are trying to compare here, but a quick simulation implies that the Purifi can do 3dB more than the SB Acoustics at 20hz before reaching its excursion limit, while according to your table you need 28 Purifi drivers to achieve the same as 4 SB Acoustics drivers.

That does not seem to compute (at all).
 
@mcdn I haven't spent much time to understand what you are trying to compare here, but a quick simulation implies that the Purifi can do 3dB more than the SB Acoustics at 20hz before reaching its excursion limit, while according to your table you need 28 Purifi drivers to achieve the same as 4 SB Acoustics drivers.

That does not seem to compute (at all).


That's what I thought, but the Purifi is a spectacularly insensitive driver at 20Hz: 67dB from the Purifi data sheet, and confirmed by hificompass measurements. The SB is 78dB sensitive at 20Hz, confirmed by AudioXpress measurements, so requires ~4x fewer drivers to reach the same SPL on that basis alone.

I haven't considered power or excursion at all, because the second half of the table uses the measured SPL to extrapolate driver numbers to achieve 110dB in-room under the measurement conditions. By definition this means each driver would be performing as measured in that scenario. Of course you can use fewer purifi drivers to get to 110dB @ 20Hz as they have huge excursion, but the distortion would be bad for the money you paid, and the amps would be huge.

There may also be a formula error with the sheet, but I suspect it's more a case of accumulated error across various approximate fields, plus a slightly nonlinear relationship between voltage and SPL for the Purifi at 20Hz.

Whether it's 4x or 7x the drivers, I was still very surprised, and happily expect to be shown I'm wrong! I'll publish a link to the actual excel sheet when I get a chance.
 
That's what I thought, but the Purifi is a spectacularly insensitive driver at 20Hz: 67dB from the Purifi data sheet, and confirmed by hificompass measurements. The SB is 78dB sensitive at 20Hz, confirmed by AudioXpress measurements, so requires ~4x fewer drivers to reach the same SPL on that basis alone.

The Purifi probably isn't the right choice since it's so expensive, but I think you're getting all this wrong. 4x fewer drivers is only if assuming the same amplifier power, and I still don't think it's right either. The SB Acoustics vs Purifi sensitivity measures seem to imply about the same loss from 50-100hz down to 20hz. The Hifi compass simulation in a 30l closed box is only 3dB down at 20hz.

I am also not sure what you mean by "Of course you can use fewer purifi drivers to get to 110dB @ 20Hz as they have huge excursion"

SPL is just about moving air, so excursion vs cone area. The Purifi has a 360cm^2 cone area, while the SB Acoustics is presented to have a 312cm^2 cone area.

So with the same amount of excursion, the Purifi will play louder. While your table seem to imply you need fewer SB Acoustics to play as loud, while it appears to be the other way around in reality. At a given excursion, the Purifi is likely to have less distortion than the SB Acoustics driver.

Finally I think it's the wrong move to look only at 20hz as a measure of performance. You are likely to play and hear 95% more 30hz.
 
That's what I thought, but the Purifi is a spectacularly insensitive driver at 20Hz: 67dB from the Purifi data sheet, and confirmed by hificompass measurements. The SB is 78dB sensitive at 20Hz, confirmed by AudioXpress measurements, so requires ~4x fewer drivers to reach the same SPL on that basis alone.

I haven't considered power or excursion at all, because the second half of the table uses the measured SPL to extrapolate driver numbers to achieve 110dB in-room under the measurement conditions. By definition this means each driver would be performing as measured in that scenario. Of course you can use fewer purifi drivers to get to 110dB @ 20Hz as they have huge excursion, but the distortion would be bad for the money you paid, and the amps would be huge.

There may also be a formula error with the sheet, but I suspect it's more a case of accumulated error across various approximate fields, plus a slightly nonlinear relationship between voltage and SPL for the Purifi at 20Hz.

Whether it's 4x or 7x the drivers, I was still very surprised, and happily expect to be shown I'm wrong! I'll publish a link to the actual excel sheet when I get a chance.
I am not understanding what you are doing.

Why don't you load the measured Thiele-Small parameters into a speaker simulator?

Here is the shallow SB acoustics sub compared to a Seas and a PuriFi sub, in sealed boxes with Q = 0.707. The Seas is one of my gold-standard subs, but the price has gone up over 2x so is now slightly ridiculous. Each driver is simulated at 250W which is close to the peak excursion for each of these subs. I included 2x SB shallow subs just because they are limited in actual output capability compared to the full-sized subs, the second sub just adds 6dB of output over the single unit. Shallow drivers have tradeoffs, the SB have non-linear motor due to the limitations in the geometry of the magnet and voice coil and are quite limited in excursion as a result, and end up being costly due to these design aspects. And the SB needs to go into a very small box, which for your situation may be an advantage.

1727192435565.png

You need ~2x SB acoustics shallow subs to get roughly the same output of one Seas L26ROY. Not a bad result. Although the non-linearities in the SB motor will likely somewhat elevate the distortion, but I have a hard time believing you will hear the difference unless you overdrive the subs.

The PuriFi is an expensive way to generate bass. Are you considering the PuriFi because of the allegedly low distortion that they advertise? I think you are overtheorizing on the distortion part of your requirements.

This sim was done on a freeware speaker simulation tool (UniBox). I recommend you use something similar. I think you are misdirecting yourself with the above table, but perhaps I am just not following.

If a shallow sub is really the only thing that fits your space, then you should get shallow subs and not worry too much. Subs are so unobtrusive I have never needed to get shallow drivers.
 
I did a quick simulation too.

SB Acoustics 20hz@90dB in a closed enclosure of 20l: 9mm excursion and requires ~90W
Purifi 20hz@90dB in a closed enclosure of 20l: 8.2mm excursion and requires ~90W

So pretty similar but less excursion and very likely less distortion with Purifi. But also a way more expensive driver.
 
If I may add the Lavoce to the equation merely for fun

In a well-damped 20l sealed box I simulated 87.7dB at 20Hz at 90W using one driver. This will get to 50.6% of the linear xmax (I guess that would mean literally no distortion in real-life)
When driven to 100% of linear xmax (approx. 350W) it will yield 93.6dB at 20Hz

90W
1727207535321.png

1727207553997.png


350W
1727207656721.png

1727207677857.png
 
  • Like
Reactions: MAB
If I may add the Lavoce to the equation merely for fun

In a well-damped 20l sealed box I simulated 87.7dB at 20Hz at 90W using one driver. This will get to 50.6% of the linear xmax (I guess that would mean literally no distortion in real-life)
When driven to 100% of linear xmax (approx. 350W) it will yield 93.6dB at 20Hz

90W
View attachment 394644
View attachment 394645

350W
View attachment 394646
View attachment 394647
Yeah, that LaVoce WAN102.50LD looks like another solid choice, and very shallow. Some of the pro-sound drivers with neo magnets high efficiency, and lower excursion are quite shallow and will work really well in sealed with DSP.
 
SPL is just about moving air, so excursion vs cone area. The Purifi has a 360cm^2 cone area, while the SB Acoustics is presented to have a 312cm^2 cone area.
Of course, this is always true (although the SB spec is low according to Vance Dickason, he measured more like 330cm^2).

What I'm looking at is something a bit different, which is SPL achievable for a given distortion level. In this case we can't use Xmax in our simulations. Since distortion and excursion are so closely related, let's invent a new parameter Xdist(f,d), which represents the maximum excursion a driver can achieve at frequency f and distortion < d. Whether it's sensible or not, I've chosen f=20 and d=10.
 
Of course, this is always true (although the SB spec is low according to Vance Dickason, he measured more like 330cm^2).

What I'm looking at is something a bit different, which is SPL achievable for a given distortion level. In this case we can't use Xmax in our simulations. Since distortion and excursion are so closely related, let's invent a new parameter Xdist(f,d), which represents the maximum excursion a driver can achieve at frequency f and distortion < d. Whether it's sensible or not, I've chosen f=20 and d=10.
I think you are looking for a set of THD vs. SPL vs. frequency contours.
For instance, this sealed 12" slim-line subwoofer can do 18% distortion at 40V input at 20Hz:
1727228472438.png


And from the CEA curves, you can see that 40V is 95dB:
1727228532143.png


It seems like you are asking for this type of data... If so, this gives a pretty good idea of what a competent 12" sealed sub can do. They make a dual 18" that can do 12% at 20Hz at 100dB.
 
Of course, this is always true (although the SB spec is low according to Vance Dickason, he measured more like 330cm^2).

What I'm looking at is something a bit different, which is SPL achievable for a given distortion level. In this case we can't use Xmax in our simulations. Since distortion and excursion are so closely related, let's invent a new parameter Xdist(f,d), which represents the maximum excursion a driver can achieve at frequency f and distortion < d. Whether it's sensible or not, I've chosen f=20 and d=10.

But that is typically what Xmax is. While not standardised, the Xmax typically implies the maximum excursion while keeping a linear response - where the definition of linear may differ from manufacturer from manufacturer. But for instance 10% THD is a common figure.

And it is reasonable to assume that a driver that can reach a given SPL at a lower excursion (in this case the Purifi), will also have lower distortion.

If I understand your current table + goals correctly, you have calculated that 4 (four) SB Acoustics will achieve 20hz at the same distortion as 28 (twenty eight) Purifi drivers. I'm just trying to tell you that this appears to be very wrong.
 
But that is typically what Xmax is. While not standardised, the Xmax typically implies the maximum excursion while keeping a linear response - where the definition of linear may differ from manufacturer from manufacturer. But for instance 10% THD is a common figure.
@sigbergaudio thank you for continuing to discuss this. I think the phrase "knows just enough to be dangerous" applies to me. The way I was thinking about the issue and using the published measurements does indeed lead to absurd results. My approach is overcomplicated, and this is compounded by the difficulties in comparing measurements between reviewers.

Reframing the problem based on your feedback, my question becomes "What is the excursion for driver X for 10% THD?". SPL can be calculated from that and Sd. I'd also forgotten about James Larsen's excellent page on the topic at https://www.audioholics.com/loudspeaker-design/audibility-of-distortion-at-bass, which goes some way toward justifying my wish for 10% THD, and mostly 2nd if possible

Upon closer reading of the Vance Dickason's measurements at AudioXpress, he is consistently using the Klippel default measure of 20% distortion to define linear Xmax for subwoofer drivers where he has Klippel analysis. The SB driver hits 20% at 8mm limited by surround stiffness, not Bl, so we can guess that it might be 10% at 5-6mm, which is way below the published Xmax of 12mm.

I still think there is a valid question here, but answering it seems very hard.
 
Back
Top Bottom