The past two weeks have been a black hole of DIY subwoofer obsession. With high-end audio gear costing more than a used car these days, I’ve started thinking—why not build my own
I’m toying with the idea of upgrading to a set of identical DIY subs in near future may be in a year. Is it a great idea or a potential money pit? No clue. But after walking through endless forum posts and YouTube rabbit holes (mostly featuring folks who think Dayton is the only driver on Earth), I’m officially ready to consult the real gurus—you guys.
The bass battleground:
13.5 ft wide x 21 ft long x 10 ft high
Its a Acoustically Treated Room (Thanks to Anthony Grimani Videos), I have M&K S150 LCRs, SVS Bookshelf Surround and Subwoofer i have two PB 2000 (Not Pro) at rear and single Rythmik FVX 15 in front
Here’s my current state of mind:
Oh, and fun fact—ChatGPT’s been throwing out suggestions like it’s got a subwoofer PhD. Honestly, it's like talking to an audio-obsessed buddy who never sleeps. 😄
I’m toying with the idea of upgrading to a set of identical DIY subs in near future may be in a year. Is it a great idea or a potential money pit? No clue. But after walking through endless forum posts and YouTube rabbit holes (mostly featuring folks who think Dayton is the only driver on Earth), I’m officially ready to consult the real gurus—you guys.
The bass battleground:
13.5 ft wide x 21 ft long x 10 ft high
Its a Acoustically Treated Room (Thanks to Anthony Grimani Videos), I have M&K S150 LCRs, SVS Bookshelf Surround and Subwoofer i have two PB 2000 (Not Pro) at rear and single Rythmik FVX 15 in front
Here’s my current state of mind:
- DIY subs: brilliant or bonkers?
- My woodworking skills are… let’s just say Sharpening the pencil .
- What good drivers are available in India? Dayton is everywhere, but I’ve got my eyes on Lavoce and BMS
- 18-inch vs. 21-inch: is bigger really better?
- WinISD is fun until it throws a “division by zero” tantrum. Can I really trust it to model things down to 20Hz?
- Driver shortlist so far: Dayton UM1,Lavoce SAN214.50 (Refered fellow forum mate!) & BMS 18N862
- Looks BMS is good.
- Sealed or Ported which one to go.
- And for designing Sealed do i need to depend Win ISD or i can go with the capacity details provided in OEM Site BMS says 150 Liter for Sealed Enclosure.
Oh, and fun fact—ChatGPT’s been throwing out suggestions like it’s got a subwoofer PhD. Honestly, it's like talking to an audio-obsessed buddy who never sleeps. 😄
Money can be saved if you have the tools and skills and don't count your time at professional cost rates.DIY subs: brilliant or bonkers?
My woodworking skills are… let’s just say Sharpening the pencil .
You can build to fit your particular aesthetics and space, so it's possible to get exactly what you want.
More displacement =more output potential.18-inch vs. 21-inch: is bigger really better?
Displacement is Sd x excursion.
More displacement costs more whether achieved with larger Sd of more excursion.
The larger the cone, the heavier it must be to withstand displacement pressure, and the more magnetic strength required to control the weight.
18" tend to be a bit more cost-effective than 21", and in general do better in smaller cabinets than 21".
Yes, you can trust it when not malfunctioning.
- WinISD is fun until it throws a “division by zero” tantrum. Can I really trust it to model things down to 20Hz?
Hornresp gives more simulation options.
1)Most "chest punch" is in the 60-160Hz range.Goal: cinematic bass that punches you in the chest—just like the legendary JBL 4645C.
- Driver shortlist so far: Dayton UM1,Lavoce SAN214.50 (Refered fellow forum mate!) & BMS 18N862
- Sealed or Ported which one to go.
- And for designing Sealed do i need to depend Win ISD or i can go with the capacity details provided in OEM Site BMS says 150 Liter for Sealed Enclosure.
Not hard to beat the JBL 4645C ( 225L net, 25Hz Fb) using the 2242H, with only 9mm Xmax (linear excursion) using modern drivers.
Doubling excursion allows +6dB more output using four times the power.
The BMS 18N862 has 19mm excursion.
https://data-bass.com/systems/5c48bfc611126b0004ca12eb
The CEA-2010 test measures the (relatively) clean output of speakers, so they can be compared "apples to apples".
The BMS 18N862 has 12dB more output at 20Hz, +4dB at 25Hz in a smaller cabinet (also 25Hz Fb) than the 4645C.
At 1kHz +10dB sounds "twice as loud", at 20Hz only ~+4dB is required to sound twice as loud.
Note the 18N862 matches or exceeds the output of the B&C21" below 50Hz in a 175L (gross) sealed box.
2) Sealed has better transient response, but at the expense of ~-6dB output compared to bass reflex in much of the lower bass around Fb.
In a small room like yours, the room modes will mask most of the difference in transient response.
3) Bigger sealed boxes require less power for the driver to reach Xmax, and require less EQ to flatten response. Volume reduction past 150L for the BMS 18N862 would probably use too much power to reach Xmax.
Art
Last edited:
Thanks a ton for the detailed reply! I'm leaning towards going with the BMS 18N862 in a sealed box. Like I mentioned earlier, my home theater setup already has 1 Rythmik FVX15 and 2 PB2000s.
Planning to add 2 sealed BMS subs for now and try to blend them with the PB2000s using EQ. Eventually, I might go all-in with 4 BMS units. The main hurdle is finding a skilled carpenter here in India.
Also, that Data-Bass link you shared has a compact vented design—looks interesting. But I keep hearing that ported boxes can be tricky to build right unless the carpenter really knows what they’re doing. So sealed feels safer for now.
Planning to add 2 sealed BMS subs for now and try to blend them with the PB2000s using EQ. Eventually, I might go all-in with 4 BMS units. The main hurdle is finding a skilled carpenter here in India.
Also, that Data-Bass link you shared has a compact vented design—looks interesting. But I keep hearing that ported boxes can be tricky to build right unless the carpenter really knows what they’re doing. So sealed feels safer for now.
I've used VituixCAD to perform a simulation of the BMS 18N826 in a 180-litre sealed enclosure. The response curve is shown below. The F3 of the system is 19.2Hz, with peak Xmax achieved at around 20Hz. The maximum SPL in the passband is around 109dB. The response filtering includes:The BMS 18N862 has 12dB more output at 20Hz, +4dB at 25Hz in a smaller cabinet (also 25Hz Fb) than the 4645C.
- 4th-order 80-Hz Linkwitz–Riley low-pass filter.
- A shelf filter with –5.5dB of cut at high frequencies.
- Parametric EQ of +6dB at 21Hz with a Q=2.0.
WinISD has the personality of a caffeinated raccoon. Use OEM specs as a baseline. Don’t trust just the “recommended enclosure” blindly, model it.The past two weeks have been a black hole of DIY subwoofer obsession. With high-end audio gear costing more than a used car these days, I’ve started thinking—why not build my own
I’m toying with the idea of upgrading to a set of identical DIY subs in near future may be in a year. Is it a great idea or a potential money pit? No clue. But after walking through endless forum posts and YouTube rabbit holes (mostly featuring folks who think Dayton is the only driver on Earth), I’m officially ready to consult the real gurus—you guys.
The bass battleground:
13.5 ft wide x 21 ft long x 10 ft high
Its a Acoustically Treated Room (Thanks to Anthony Grimani Videos), I have M&K S150 LCRs, SVS Bookshelf Surround and Subwoofer i have two PB 2000 (Not Pro) at rear and single Rythmik FVX 15 in front
Here’s my current state of mind:
Goal: cinematic bass that punches you in the chest—just like the legendary JBL 4645C.
- DIY subs: brilliant or bonkers?
- My woodworking skills are… let’s just say Sharpening the pencil .
- What good drivers are available in India? Dayton is everywhere, but I’ve got my eyes on Lavoce and BMS
- 18-inch vs. 21-inch: is bigger really better?
- WinISD is fun until it throws a “division by zero” tantrum. Can I really trust it to model things down to 20Hz?
- Driver shortlist so far: Dayton UM1,Lavoce SAN214.50 (Refered fellow forum mate!) & BMS 18N862
- Looks BMS is good.
- Sealed or Ported which one to go.
- And for designing Sealed do i need to depend Win ISD or i can go with the capacity details provided in OEM Site BMS says 150 Liter for Sealed Enclosure.
Oh, and fun fact—ChatGPT’s been throwing out suggestions like it’s got a subwoofer PhD. Honestly, it's like talking to an audio-obsessed buddy who never sleeps. 😄
Get the board laser cut, then assemble.
Not much need of carpenter in that case.
Not much need of carpenter in that case.
Since when is a carpenter supposed to design a ported box? In my experience, a carpenter builds what you have designed.ported boxes can be tricky to build right unless the carpenter really knows what they’re doing.
You are right.. its making me crazy 🙂 btw the following is my first try to design the Sealed box. Also if i try to add LT filter the cone excursion goes crazy with 800 watts. I am doing something wrong ? is it ok add the filter and check ? and is120 liters gross volume fine to achieve 0.707 QTC ?
Thank you ! i will try the option do you have any reference in India ? even for sourcing the driver ? i tried the distributor mentioned in BMS site. but no response from them.Get the board laser cut, then assemble.
Not much need of carpenter in that case.
With the chosen LT filter design, about +7.0dB of boost is being applied at 30Hz. This increases to +8.5dB at 20Hz or so. That's a lot of boost, so little wonder the cone excursion "goes crazy with 800 watts"! Adding the LT filter sends a lot of extra power to the driver.Also if i try to add LT filter the cone excursion goes crazy with 800 watts. I am doing something wrong ?
By the way, I can't see the full width of the plots that you have included in your post. You may need to adjust the manner in which they are included.
No, that gross volume is too large. The Qtc will be around 0.62, so the low-frequency response is still quite damped. If filling material is added, then the effective volume will be greater, maybe up around 150 litres or so. That will further reduce the Qtc to 0.58 or so. To get to Qtc=0.71, the effective enclosure volume needs to be around 81 litres. That will push the F3 point to about 54Hz or so.is it ok add the filter and check ? and is120 liters gross volume fine to achieve 0.707 QTC ?
If we add a shelf filter with +6dB of boost at low frequencies, we can achieve the following response. Is that something that might suits your purposes? It won't have the extreme excrusions at low frequencies that were introduced by the previously chosen LT filter.
Most big cities have laser cutting services for steel and wood, quite common now.
Ask the local board supplier, or get it cut by them on a table saw...not expensive at all.
Drivers...many suppliers, ask at local shops, it is a cottage industry, you can ask local speaker builders or dealers.
I know at least two makers here in Baroda, 16" driver was around 2500 Rupees two years back at the factory.
Copper has gone expensive after that...
There are many sellers in Delhi and Bombay, YMMV, I do not guarantee their work.
Ask the local board supplier, or get it cut by them on a table saw...not expensive at all.
Drivers...many suppliers, ask at local shops, it is a cottage industry, you can ask local speaker builders or dealers.
I know at least two makers here in Baroda, 16" driver was around 2500 Rupees two years back at the factory.
Copper has gone expensive after that...
There are many sellers in Delhi and Bombay, YMMV, I do not guarantee their work.
Thank you !! 🙂With the chosen LT filter design, about +7.0dB of boost is being applied at 30Hz. This increases to +8.5dB at 20Hz or so. That's a lot of boost, so little wonder the cone excursion "goes crazy with 800 watts"! Adding the LT filter sends a lot of extra power to the driver.
By the way, I can't see the full width of the plots that you have included in your post. You may need to adjust the manner in which they are included.
No, that gross volume is too large. The Qtc will be around 0.62, so the low-frequency response is still quite damped. If filling material is added, then the effective volume will be greater, maybe up around 150 litres or so. That will further reduce the Qtc to 0.58 or so. To get to Qtc=0.71, the effective enclosure volume needs to be around 81 litres. That will push the F3 point to about 54Hz or so.
If we add a shelf filter with +6dB of boost at low frequencies, we can achieve the following response. Is that something that might suits your purposes? It won't have the extreme excrusions at low frequencies that were introduced by the previously chosen LT filter.
===============
I have reduced the net volume qtc 0.706.
And adding filters not helping for me 🙁 could you please tell me how to add in Winisd ? i will also explore..
Also, is it fine to add MLP distancer in signal tab ?
- Home
- Loudspeakers
- Subwoofers
- Help DIY Subwoofers - India