Preface
I have a 2-way sealed speakers with 10" midbass, which is better suited for midrange and took a lot of DSPing to get good response down to 50 Hz. I cross to sealed 14" subs at 100 Hz when I want to. I was wondering if I could have better bass quality, not just SPL, by adding a woofer for 50~200 Hz range (presumption: yes). I'm using 3-way DSP plates in existing speakers to drive the midbass in BTL-mode, but I could have the third channel to drive a separate woofer instead. I have some conditions: 1) I was hoping a relatively small box (less than 50 liters) to remain stand-mounted (on the subs) and low floor footprint. If I have to go floor-standing, the solution would probably need to be able to ditch the subs. 2) I hate port noise and muffled response below port tuning frequency, so sealed it is. As are my current main speakers and subs.
Beef
I've read through many forum threads and hundreds of 12-15" woofer specs (not joking), and started modeling with VituixCAD. I noticed that virtually any 15" is easy to DSP to 50 Hz or lower response, even with way too small box and completely ignore any EBP and Q-factors. With 12" the box size probably fits my target box size external limitations, but takes a bit more effort to get to 50 Hz. My current 10" woofers have a fairly large box for sealed alignment, so this experience does not directly translate to sealed in a small box.
The real questions are:
1) How to choose a good-sounding woofer for this frequency range, which T/S parameters are meaningful to predict transient quality etc?
2) Will a small, sealed box hurt woofer sound quality, if it had T/S params more suited for bass reflex operation (like high BL, Qts<0.3, high VAS and so on)? This essentially means 12" vs. 15" in my case.
3) Should I just aim for 20 Hz with a 15" or an 18" in a proper box and ditch the subs?
A funny thing is that I currently have a 3-way system as a whole and would be just trading the subs for woofers and cross higher. But I don't think that would be a downgrade and also going 4-way system as a whole is tempting. The full stack could get quite tall, but I can go creative with element placement when they are all in different boxes. If the prototype fulfills expectations, I'll probably make new cabinets to house all three divers.
PS
My room is largish and quite free from bass-modes. I'm using DSP-plate amps and a DSP-preamp, so almost any form and amount of DSP is a possibility. I also have good amps lying around that I could use for power. I'm fond of Eighteen Sound products and hope to find a solution there.
I have a 2-way sealed speakers with 10" midbass, which is better suited for midrange and took a lot of DSPing to get good response down to 50 Hz. I cross to sealed 14" subs at 100 Hz when I want to. I was wondering if I could have better bass quality, not just SPL, by adding a woofer for 50~200 Hz range (presumption: yes). I'm using 3-way DSP plates in existing speakers to drive the midbass in BTL-mode, but I could have the third channel to drive a separate woofer instead. I have some conditions: 1) I was hoping a relatively small box (less than 50 liters) to remain stand-mounted (on the subs) and low floor footprint. If I have to go floor-standing, the solution would probably need to be able to ditch the subs. 2) I hate port noise and muffled response below port tuning frequency, so sealed it is. As are my current main speakers and subs.
Beef
I've read through many forum threads and hundreds of 12-15" woofer specs (not joking), and started modeling with VituixCAD. I noticed that virtually any 15" is easy to DSP to 50 Hz or lower response, even with way too small box and completely ignore any EBP and Q-factors. With 12" the box size probably fits my target box size external limitations, but takes a bit more effort to get to 50 Hz. My current 10" woofers have a fairly large box for sealed alignment, so this experience does not directly translate to sealed in a small box.
The real questions are:
1) How to choose a good-sounding woofer for this frequency range, which T/S parameters are meaningful to predict transient quality etc?
2) Will a small, sealed box hurt woofer sound quality, if it had T/S params more suited for bass reflex operation (like high BL, Qts<0.3, high VAS and so on)? This essentially means 12" vs. 15" in my case.
3) Should I just aim for 20 Hz with a 15" or an 18" in a proper box and ditch the subs?
A funny thing is that I currently have a 3-way system as a whole and would be just trading the subs for woofers and cross higher. But I don't think that would be a downgrade and also going 4-way system as a whole is tempting. The full stack could get quite tall, but I can go creative with element placement when they are all in different boxes. If the prototype fulfills expectations, I'll probably make new cabinets to house all three divers.
PS
My room is largish and quite free from bass-modes. I'm using DSP-plate amps and a DSP-preamp, so almost any form and amount of DSP is a possibility. I also have good amps lying around that I could use for power. I'm fond of Eighteen Sound products and hope to find a solution there.
Last edited:
It's all about finding the woofer with sufficient displacement. and the power handling to move the cone that far. You start with how loud you want it to play the lowest frequency. So maybe you want 30 Hz at an SPL at 1 meter to be 102 dB. You can go to the linkwitz web page ( https://www.linkwitzlab.com/thor-estim.htm ) and get the spreadsheet to calculate the required displacement. A 12" woofer with 0 to peak Xmax of 12.5 mm and cone area 466 cm^2 will do that. Then you find some woofers with those parameters and try a few different configurations in WinISD to see that the driver in the box size you want will be able to reach Xmax withing it's power limit. If the box is really small the added stiffness can make that become an issue. I usually just build the box size as recommended or a bit smaller and then use the Linkwitz Transfer to extend the bass to the desired low frequency cut off. It works great. II think WinISD has a Linkwitz transfer filter model built in that will let you model the result. Here's the equation as presented by Jeff Bagby a while back for a sealed box in half space ( on floor away from walls) Against one wall would give 6 dB? more gain. In a corner more gain still. Take woofer room position into account when designing can save a lot of money. Also look at the room simulator in Room EQ Wizard as it can show you what to expect for woofer placements.
SPL= 20 * log(10) (1.18 / 0.00002 * Sd * Xmax / SQRT(2) * 2 * PI * F^2)
Where Sd is in Meters^2
Xmax is in Meters one-way linear travel
And F is the frequency you want to solve for.
You might look at the Scan Speak Discovery woofers, as they have large Xmax and work in small boxes.
SPL= 20 * log(10) (1.18 / 0.00002 * Sd * Xmax / SQRT(2) * 2 * PI * F^2)
Where Sd is in Meters^2
Xmax is in Meters one-way linear travel
And F is the frequency you want to solve for.
You might look at the Scan Speak Discovery woofers, as they have large Xmax and work in small boxes.
Last edited:
@olson3 Thanks for the reply, but I've already simulated a good number of drivers with VituixCAD, which is a lot more advanced than WinISD. VCad has Linkwitz transform for a filter and that's what I've used to simulate in conjunction with boxes of certain sizes. It also gives much more useful information, such as excursion per given power, group delay, phase, impedance and everything you need for a full speaker.
I have yet to try VituixCAD. WinISD does all the calculations you mentioned, its all the same math, but it does not model passive crossovers. WinISD lets you directly compare multiple projects on a single graph, where each project could have different woofer and or box size etc. Once you decide on the Max SPL level you want at a low frequency you can use the calculation I showed above to determine the required displacement. From there you can easily determine which woofers have sufficient displacement, and you're done. At low frequencies all these drivers are operating as a piston, so there isn't a good sounding driver. Differences in transient response are mute as you are going to band limit to 200 Hz. What SPL at what low frequency limit are you targeting?
As already mentioned, first try to work out the in-seat SPL and then use simulators to arrive at a proper driver with the correct excursion and power handling.
Do have a look at the amplifier VA at full power and excursion, after the EQs are applied.
Drivers with comparable efficiency, Xmax and power handling would sound more or less the same in this frequency range as you don't hear much sound, but just feel some vibrations. Transient performance is already at best for sealed boxes. EBP is what you need to look at, for sealed / vented decisions.
The sealed / vented suggestion is only a suggestion and actual sound quality depends on SPL / power / excursion, distortion, power compression etc. The 15" driver would make bass more easily when compared to the 12", as it has more cone area that can move a higher volume of air per unit excursion.
Definitely not, but get better (18" or 21") subwoofer for this range to make bass more easily. These frequencies when mixed up with regular sound could result in intermodulation distortions, so it's best to keep them coming from a separate driver.
Most 3-ways are actually 4-ways and it's just that people keep forgetting the subs, so it's not uncommon. Most subs go somewhere else (in-between, corner etc.) and not right under the main speakers. That way, the boundary reinforcement for the mains are also maximised.
I would throw subwoofers away and use 15" or 18" subbass to midbass range. Something like ciare NDH15-4S would work. (Rare pro woofer, low fs and high xmax)
That NDH15-4S would work nicely. In a half cube box the first box mode can be around 700 Hz as shown here. Easy to control with a bit of stuffing. http://www.loudspeakerdatabase.com/...=20/lmz=1/lmrf=20000/vr=0.0/nm_max=14/tbox=22
Here it is with 50 mm of damping: http://www.loudspeakerdatabase.com/...20/lmz=50/lmrf=20000/vr=0.0/nm_max=14/tbox=22
Attachments
Keep a lookout for Peerless XXLS-P830845.
I acquired mine new last year, and it was manufactured the same year.
But depending on the market might be challenging to source, however this is still active Tymphany product and in their catalogue:
https://products.peerless-audio.com/transducer/207
Old-fashioned combination of good sensitivity moderate sealed box size and bass extension. 8 ohm nominal. Oldie but a goodie and 13mm Klippel verified x-max by user npdang on DIYMobileAudio forums years ago.
Usable to 300Hz IMHO (first resonance at 600Hz). Model it in 2cu feet ea woofer in VituixCAD with baffle step loss and see if it meets your needs. The auto-align method asks for a ridiculously large cabinet, so you might have missed this woofer in your shortlist.
I run 2 in parallel for 92-93 dB/2.83V (4 ohm) in a slim cabinet that most commercial manufacturers put a 8-10" woofer in. eg Revel F328Be (340mm wide, 450mm deep) F3 40Hz anechoic, IIRC. I have DSP but in my room I don't even need to dial it in...
I hate flowery adjectives too, but slamming tight bass is what it is. Perfectly excursion AND thermally limited at around 400W. It's the biggest speakers in this post, next to the TV. I use it as my bass target when I design smaller speakers.
https://www.diyaudio.com/community/...ts-purifi-and-scan-speak.390316/#post-7148763
It really is a winner in my book, as long as you can get your hands on it...
I acquired mine new last year, and it was manufactured the same year.
But depending on the market might be challenging to source, however this is still active Tymphany product and in their catalogue:
https://products.peerless-audio.com/transducer/207
Old-fashioned combination of good sensitivity moderate sealed box size and bass extension. 8 ohm nominal. Oldie but a goodie and 13mm Klippel verified x-max by user npdang on DIYMobileAudio forums years ago.
Usable to 300Hz IMHO (first resonance at 600Hz). Model it in 2cu feet ea woofer in VituixCAD with baffle step loss and see if it meets your needs. The auto-align method asks for a ridiculously large cabinet, so you might have missed this woofer in your shortlist.
I run 2 in parallel for 92-93 dB/2.83V (4 ohm) in a slim cabinet that most commercial manufacturers put a 8-10" woofer in. eg Revel F328Be (340mm wide, 450mm deep) F3 40Hz anechoic, IIRC. I have DSP but in my room I don't even need to dial it in...
I hate flowery adjectives too, but slamming tight bass is what it is. Perfectly excursion AND thermally limited at around 400W. It's the biggest speakers in this post, next to the TV. I use it as my bass target when I design smaller speakers.
https://www.diyaudio.com/community/...ts-purifi-and-scan-speak.390316/#post-7148763
It really is a winner in my book, as long as you can get your hands on it...
Last edited:
Now that you've probably guessed how Xmax is possibly the most important parameter, why don't you show some of your preferred models? Maybe you could work forward from there..
Yes, that's a good option. Great woofer. 6 ohm nominal impedance, and in pairs in parallel in 100L, the min impedance is 2.2 ohm vs 3.1 ohm for the XXLS. No problem if you're using modern class D amplifier.
It it clean up to past 500Hz?
The only fly in the ointment for the XXLS-P830845 is the breakup at ~600Hz. There's definitely a destructive dip there, just like the datasheet shows, so the absolute upper crossover limit is 600Hz IMHO, so it's no mid-woofer.
A third option for twin twelves as woofers (not just subwoofers) is Dayton Reference series RSS315HFA-8 is also a great woofer. Can play a much higher (past 500Hz), but voltage sensitivity lower is much lower. Anechoic 92dB/2.83V but only about 87dB/2.83V @ 100Hz in a real cabinet due to baffle step losses.
Last edited:
Hi,
if you play that high up with the woofer then absolutely choose a driver that is great on the upper end of the bandwidth! Subwoofers are optimized for low end, not for midrange.
Because, if the midrange is not good you have no options but to change the driver for better one so you might as well do it from the start. You can always add more bass woofers, subwoofers, for the low end if it ends up lacking. Just pour in as many and big woofers as needed, under the sofa, behind the fridge, in the closet and so on. Midrange is critical, always optimize for the upper end of the bandwidth. You could use subwoofer driver, but make sure it has every trick on it to make midrange better, shorting rings, no cone resonances, big enough Sd to keep excursion low enough to name few. Also, make the box so it doesn't ruin mids, take care of resonances.
Again, if you are after good sound, then low end response of any "way" can and should be sacrificed for better upper end results as there is always possibility to add another driver(s) for the lows. If the upper end of the bandwidth is compromised then there is no way to get the performance back other than doing it again better, another driver, another structure.
Hope it helps, have fun!🙂
if you play that high up with the woofer then absolutely choose a driver that is great on the upper end of the bandwidth! Subwoofers are optimized for low end, not for midrange.
Because, if the midrange is not good you have no options but to change the driver for better one so you might as well do it from the start. You can always add more bass woofers, subwoofers, for the low end if it ends up lacking. Just pour in as many and big woofers as needed, under the sofa, behind the fridge, in the closet and so on. Midrange is critical, always optimize for the upper end of the bandwidth. You could use subwoofer driver, but make sure it has every trick on it to make midrange better, shorting rings, no cone resonances, big enough Sd to keep excursion low enough to name few. Also, make the box so it doesn't ruin mids, take care of resonances.
Again, if you are after good sound, then low end response of any "way" can and should be sacrificed for better upper end results as there is always possibility to add another driver(s) for the lows. If the upper end of the bandwidth is compromised then there is no way to get the performance back other than doing it again better, another driver, another structure.
Hope it helps, have fun!🙂
Inner walls are so light that bass goes through quite easily and gets mostly absorbed and diffused in adjacent room. Calc predicts room modes for outer walls of the house for 34.4 Hz and 28.7 Hz, but neither is there.
Here are two 15" drivers that have quite different T/S parameters. Target EQ is -3 db to 50 Hz, but room gain has not been factored in, which would make it easier. The graphs are from VituixCAD enclosure tool, a crossover (active EQ filters) is applied individually for each.
Eighteen Sound 15W500 in a 57 litre box.
Eighteen sound 15ND930 in a 54 litre box.
The former driver is seemingly easier to EQ for low end extension. While both easily achieve 50 Hz @ 100 dB, looking at these graphs the latter is a better choice. But either would do fine in home use, so how should I reason a choice? 🙂
As tmuikku pointed out, I guess when going up to (lower) mid-range frequencies, performance there should dominate. I'll probably look for drivers with high recommended frequency range and least of resonances there.
VCad also has a warning for compression distortion (1.4 * Sd * Xmax / Vb * 100%), I've got no idea how audile it is and which percentage I should target at these frequencies, but it speaks for larger box.
Eighteen Sound 15W500 in a 57 litre box.
Eighteen sound 15ND930 in a 54 litre box.
The former driver is seemingly easier to EQ for low end extension. While both easily achieve 50 Hz @ 100 dB, looking at these graphs the latter is a better choice. But either would do fine in home use, so how should I reason a choice? 🙂
As tmuikku pointed out, I guess when going up to (lower) mid-range frequencies, performance there should dominate. I'll probably look for drivers with high recommended frequency range and least of resonances there.
VCad also has a warning for compression distortion (1.4 * Sd * Xmax / Vb * 100%), I've got no idea how audile it is and which percentage I should target at these frequencies, but it speaks for larger box.
And here goes the same 15ND930 into a 40 liter box:
Stats at 50 Hz: 99 dB, 7 Watts, excursion 1.6 mm (Xmax 7.5 mm) 😱) Power below 50 Hz is irrelevant to me, since I'm most likely going to apply high-pass filter. Is it this simple or am I missing something? Yes it would be a different story if my target was 15 Hz, but it's 50, so I nothing to worry about? 15 inch is large and efficient when you compare to 6 inchers, no matter how you EQ it, right?
Stats at 50 Hz: 99 dB, 7 Watts, excursion 1.6 mm (Xmax 7.5 mm) 😱) Power below 50 Hz is irrelevant to me, since I'm most likely going to apply high-pass filter. Is it this simple or am I missing something? Yes it would be a different story if my target was 15 Hz, but it's 50, so I nothing to worry about? 15 inch is large and efficient when you compare to 6 inchers, no matter how you EQ it, right?
Target SPL @ listening distance? The driver with lesser excursion (X/Xmax) for the required SPL / power should do better, in general. If X is not Xmax, you can equalise lower, at the expense of linearity.
Last edited:
Or play louder instead if desired.
You might also want to include that high-pass filter in your Xmax estimation, if you haven't. Same for sealed subs.
That is one of the questions in the first post. From what I've read through out many manufacturers specsheets, only a small factor of drivers are specifically tuned for closed boxes. Generally you get more features in drivers that are intended for vented boxes, eg. shorting rings, stronger motors and cooler looks. So, if my targets are modest, is there anything beyond the above graphs that I should look for?
Also the first driver with graphs, 15W500 is one of the most suited for closed boxes by EBP from all drivers I've read about, but it's still performing worse (longer travel, closer to Xmax, longer GD, draws more power) than the 15ND930, so that alone won't necessarily mean anything. That got me wondering, if picking just any driver will do?