The reason I asked is because @witwald used 60l closed box for his last closed box calculations. For the Nero, WinISD indicates 44 liters for a 0.707Q closed box. The predicted slope doesn't change much from the 60l box, just a touch over 1 dB lower at 20Hz. 1 dB doesn't seem like a ton of amp power -- if I can assume power/SPL remains linear.You can do it, but it will take a lot of boost and a ton of amplifier.
They both require a lot of power to deal with 12-16db of boost needed to get to 25hz, but watts are pretty cheap these days.
In the 60l box, the total boost at 20Hz seems to be 8dB + the LT, the boost of which I can't quite figure out from the @witwald post. If it's ~12 dB total boost, it might be OK. Probably fine for me. But any more, and I might need more than the 600W max into one channel I can muster from my current amps.
I'm cutting a baffle for the woofer, roughly 19x16", with the depth of the box to determine Vb. If I make it 20" deep, it works out to >60 liters. A bit more cube-like & it gets closer to 40l. But meanwhile, I can attach the baffle to a 2' long 16" diameter sonotube which already has end caps. Vb ends up being ~75 liters, which can be adjusted by taking up volume with things like closed cell foam. Or bricks. Or... I will use it as a test box & try things experimentally.
Both driver' Fs measure pretty close to factory specs: 32.4 Hz & 32.8 vs 31. Qts is lower -- 0.288 instead of 0.34. Havne't done any break in but won't bother for now. I don't think it'll take long with big bass signals.
It's big and heavy, a bit awkward to handle but not bad. I will not be trying to cut a close fitting recess for this frame!
The frame supports look massive & solid but they're actually a 3-sided casting.
The back vent is a tunnel with an 8-spoke tunnel profile. It looks like that's the way the magnet(s?) are shaped.
This is the 2' long 16" cardboard concrete form I'd prepared with internal ribs & support beams for a SB12" woofer as shown. I will remove those internal parts. It's big enough for the Nero-15sw800.
I'm cutting a baffle for the woofer, roughly 19x16", with the depth of the box to determine Vb. If I make it 20" deep, it works out to >60 liters. A bit more cube-like & it gets closer to 40l. But meanwhile, I can attach the baffle to a 2' long 16" diameter sonotube which already has end caps. Vb ends up being ~75 liters, which can be adjusted by taking up volume with things like closed cell foam. Or bricks. Or... I will use it as a test box & try things experimentally.
Both driver' Fs measure pretty close to factory specs: 32.4 Hz & 32.8 vs 31. Qts is lower -- 0.288 instead of 0.34. Havne't done any break in but won't bother for now. I don't think it'll take long with big bass signals.
It's big and heavy, a bit awkward to handle but not bad. I will not be trying to cut a close fitting recess for this frame!
The frame supports look massive & solid but they're actually a 3-sided casting.
The back vent is a tunnel with an 8-spoke tunnel profile. It looks like that's the way the magnet(s?) are shaped.
This is the 2' long 16" cardboard concrete form I'd prepared with internal ribs & support beams for a SB12" woofer as shown. I will remove those internal parts. It's big enough for the Nero-15sw800.
Last edited:
I cranked the handle using VituixCAD to simulate the NERO-15SW800 in a closed-box enclosure whose volume is Vb=40litres. The filters and EQ that I used were as follows:Could you look at what happens in an even smaller sealed box, say 40 liters? Without the Linkwitz Transform, perhaps, just some judicious PEQ? Can it be made to go as low, say 25 Hz, and still put out at least 100dB at Xmax at that frequency?
- 4th-order Linkwitz–Riley low-pass filter set to a 70Hz cut-off frequency. This produces an acoustic −6dB point of 80Hz.
- PEQ #1: +9.0dB at 20Hz with Q=1.50.
- PEQ #2: +6.0dB at 25Hz with Q=1.00.
To get 100dB at 25Hz requires a power input of 185W, with a displacement of about 7.5mm. This is about 1/2 of Xmax=14.3mm, so we have some displacement in reserve.
The VituixCAD simulation is shown below, where the dash-dot cyan line is the woofer response before application of EQ and filtering. The EQ+filter curve is the red curve, with the dB scale on its right-hand axis.
@witwald -- thank you. 🙂 That's more boost than I'd like, but it's good to know the relationship between Vb, EQ boost, Xmax, and SPL. 🙂
I will experiment with the tube first, then build a box 50~60l. It will be interesting, too, to compare the tube vs the box for bass quality.
Can anyone comment on the audibility of group delay shown in the above graph? There must be some price to be paid for lots of EQ? The peak of 12ms is much less than the >30ms of the 100l ported projections in your first post here (#15) but no different than the last 60l closed box projection of post #27.
I will experiment with the tube first, then build a box 50~60l. It will be interesting, too, to compare the tube vs the box for bass quality.
Can anyone comment on the audibility of group delay shown in the above graph? There must be some price to be paid for lots of EQ? The peak of 12ms is much less than the >30ms of the 100l ported projections in your first post here (#15) but no different than the last 60l closed box projection of post #27.
Last edited:
Ah. I thought this could be set for various speaker positions -- ie, full space, half quarter, etc.The sim predicts anechoic space response.
In any case, that's a good reminder. Subs are usually positioned on the floor & close to the wall for convenience, so half/quarter space mostly.
It certainly is a lot of boost, and Xmax isn't even reached.That's more boost than I'd like, but it's good to know the relationship between Vb, EQ boost, Xmax, and SPL.
The peak in the red curve in the graph in post #46 was not the group delay. It was the total filter EQ.Can anyone comment on the audibility of group delay shown in the above graph? There must be some price to be paid for lots of EQ? The peak of 12ms is much less than the >30ms of the 100l ported projections in your first post here (#15) but no different than the last 60l closed box projection of post #27.
I've added the group delay to the traces. I've also adjusted the low-pass filter to be at 120Hz. This matches its setting in the earlier simulations of the vented-box subwoofer. The new results are shown below. As you can see, the peak group delay for the closed-box subwoofer is 31ms @ 19.7Hz, similar to that for the vented-box subwoofer, although the shapes of the group delay curves are a bit different.
If we increase the −3dB low-frequency cut-off point from 20.3Hz to 24.8Hz, we can get a higher maximum SPL for a given power input. For the previous peak power input of 250W, we now achieve 104.4dB in the passband, rather than 101.3dB, which is a 4.1dB improvement. There is also a reduction in the group delay, which results from the somewhat less-extended low-frequency response.
The filters and EQ that were used above were as follows:
The filters and EQ that were used above were as follows:
- 4th-order Linkwitz–Riley low-pass filter set to a 120Hz cut-off frequency.
- PEQ #1: +6.5dB at 22Hz with Q=1.30.
- PEQ #2: +8.0dB at 29Hz with Q=1.30.
That's an interesting question.There must be some price to be paid for lots of EQ?
As we are dealing with a minimum-phase system, whatever the computed bandpass response function is, it simply has its intrinsic group delay.
For a closed-box subwoofer, the roll-off rate below its cut-off frequency is 12dB/octave. However, when the two sets of heavy EQ are applied in the rolling-off region, the local low-frequency roll-off rate has been increased to around 20dB/octave. This is almost as high as the roll-off rate that occurs when a vented-box subwoofer is used, and of course the group delay characteristics would also be similar. Note that there is an added degree of difference because the applied EQ has also resulted in a sharper knee in the magnitude response function than would typically be present.
The simulations are computed for an anechoic loading. Placing the woofer close to one or more room boundaries will of course serve to increase its output. There has also been no attempt to try and simulate the effects of room gain. In that way, in terms of maximum SPL output, the in-room performance of the subwoofer will be higher than the figures computed by the simulation. And, of course, room modes will affect the low-frequency response in the region where the subwoofer will be active, and they have not been considered either.
I used the 15sw800 for a home subwoofer. I'm pretty new to speaker building, but I modelled a basic L-slot box for it tuned to ~30Hz. I like the smooth sound and puts out decent volume (I sacrificed some dB for form factor). I run it with a bridged XLS1502, but rarely use more than the first indicator light of power at home. I only have home measurements, which are very affected the room it's in, but one day I'll get try to get it it outside to measure and post up data if you're interested.
Attachments
Busy run of days.... but some progress was made on this project.
After playing around with the temporary setup with the sonotube shown earlier, I determined with measurements & listening:
1. Qts = 0.54, Fs=54Hz. Vb=2.8cf
2. The SBA 15SW800 in this sealed volume, with the EQ suggested by @witwald in post 51 is roughly equal to the 2 pairs of SBA SB29SWNRX-S75 in W-config OB that are in my LX521 clones. Running an LR4 at 60~80Hz between the LX521 bass bin and the sealed EQ'd 15SW800 provides little or no gain. There might be the sense of a touch more impact due to room pressurization, but the original LX521 bass is excellent anyway.
3. Running the EQ'd 15SW800 in the sonotube in parallel with the LX521, with low pass LR4 at 60Hz makes a difference when there is deep bass material or ambient cues in the low frequencies -- such as recordings in big churches, orchestral pieces in large halls, or some live jazz/pop recordings. The weight of low bass brings another level of realism that's easily heard by me (and a couple friends).
So, I chose to build a box the same 2.8cf internal volume as the sonotube out of 18mm BB plywood. Panels & most of the internal braces have been cut, shaped, routered, etc. The usual mix of minor fine adjustments due to less than perfect cuts, etc....
Four more braces will be cut & installed to support the front panel & join it to the internal braces. Another 15mm panel will be glued to the front baffle in lieu of a recess/rebate for the driver, so the baffle will end up 33mm thick. The back of the magnet will be pressed up firmly against the brace behind it when the driver is bolted down. Old simple idea of providing a "sink" for vibrations from the driver frame to be dissipated into the box (instead of going into the cone). Not sure if the box will be heavy enough. I'm guessing it might get to 60~70 lbs. It's most of a full 5x5' 18mm BB sheet.
Nothing is glued yet, just a few screws on the bottom.
Another day or 2 and it should be cooking. 😉
After playing around with the temporary setup with the sonotube shown earlier, I determined with measurements & listening:
1. Qts = 0.54, Fs=54Hz. Vb=2.8cf
2. The SBA 15SW800 in this sealed volume, with the EQ suggested by @witwald in post 51 is roughly equal to the 2 pairs of SBA SB29SWNRX-S75 in W-config OB that are in my LX521 clones. Running an LR4 at 60~80Hz between the LX521 bass bin and the sealed EQ'd 15SW800 provides little or no gain. There might be the sense of a touch more impact due to room pressurization, but the original LX521 bass is excellent anyway.
3. Running the EQ'd 15SW800 in the sonotube in parallel with the LX521, with low pass LR4 at 60Hz makes a difference when there is deep bass material or ambient cues in the low frequencies -- such as recordings in big churches, orchestral pieces in large halls, or some live jazz/pop recordings. The weight of low bass brings another level of realism that's easily heard by me (and a couple friends).
So, I chose to build a box the same 2.8cf internal volume as the sonotube out of 18mm BB plywood. Panels & most of the internal braces have been cut, shaped, routered, etc. The usual mix of minor fine adjustments due to less than perfect cuts, etc....
Four more braces will be cut & installed to support the front panel & join it to the internal braces. Another 15mm panel will be glued to the front baffle in lieu of a recess/rebate for the driver, so the baffle will end up 33mm thick. The back of the magnet will be pressed up firmly against the brace behind it when the driver is bolted down. Old simple idea of providing a "sink" for vibrations from the driver frame to be dissipated into the box (instead of going into the cone). Not sure if the box will be heavy enough. I'm guessing it might get to 60~70 lbs. It's most of a full 5x5' 18mm BB sheet.
Nothing is glued yet, just a few screws on the bottom.
Another day or 2 and it should be cooking. 😉
Regarding acoustic stuffing a sub that will likely never be crossed higher than... 120Hz, what do sub gurus recommend & how much? I have high density stuff (blue fill - cotton), pillow stuffing, and lambs wool.
The last 4 braces got made & fitted this morning. Two of the 4 have ledges upon which the back of the woofer rest. They are ~0.5mm too tall, so when the woofer is bolted in place, the small contact wood surface will get crushed a bit for a tight mechanical bond. There will be 2 more wood spacers wedged in to balance the pressure around the back rim of the woofer.
All the internal braces have rounded edges -- run with a quarter-round bit on a router. Likely little/no difference but peace of mind that turbulence isn't a problem in there.
It's going to be glued/screwed later today.
I'm not sure what the function of those 8 holes on the back cover is, but incase they're for heat dissipation, I am making sure they are not blocked.
All the internal braces have rounded edges -- run with a quarter-round bit on a router. Likely little/no difference but peace of mind that turbulence isn't a problem in there.
It's going to be glued/screwed later today.
I'm not sure what the function of those 8 holes on the back cover is, but incase they're for heat dissipation, I am making sure they are not blocked.
Thanks! I thought as much -- damping not necessary. As for the carpentry... thanks. 🙂 In reality, my skills are moderate but I have lots of tools (that I mostly haven't fully mastered), I go slow, and still make lots of errors. I'm getting better at covering up or dealing with those errors.For low q sealed box not playing any higher than 120 hz I'm not sure it matters. Willing to be corrected, but I tried my Thors stuffed and unstuffed and could not tell the difference.
You have some good carpentry skills btw!
- Home
- Loudspeakers
- Subwoofers
- Help with 15" SB Audience sub project