That looks great. So, if I choose the ported design I should aim for a 28hz tune with 160 liters. How do I get this in a drawing? With subbox.pro?
Considering I 'only' have one Behringer inuke 6000dsp amp for these 2 18" drivers, will it be better if I buy the 4 ohm version and does that make a difference to these results?
Considering I 'only' have one Behringer inuke 6000dsp amp for these 2 18" drivers, will it be better if I buy the 4 ohm version and does that make a difference to these results?
The 15" Daytons should have absolutely no problem playing up to 100Hz with steep filtering. They have large amounts of shorting in the motor and subsequently have a pretty low value for Le given the size of the voice coil in them. This can also be seen in how extended their frequency response actually is.
They have a pickle of a resonance at about 330Hz and this will probably trickle down into increased third order distortion at about 110Hz but provided you've used a 4th order targeted at 100Hz this should be fine.
The main thing about bass reproduction is that it's dominated by the room and much less so by the actual speakers themselves. At least so long as the speakers (subs) aren't poorly designed.
When you're altering the crossover frequency, between the sub and your mains, you're also altering the interplay of the room modes from the various bass sources and their positions within the room. More bass sources is better. More bass sources that are spread into different locations are even better.
The higher you take the single sub the worse, and more defined, the room modes are going to be over the region that the sub is covering. That being Vs the two separate bass sources of your mains.
Changing the sub isn't going to solve this. Ideally you'd want to run the sub up to 100Hz and the mains down as low as they'll happily go. This creates an overlap between them and provides you with 3 separate bass sources. Obviously this will create an excess of output where the 3 overlap but this is simple enough to fix using some shelving filters or PEQ.
As an example sub plays up to 100Hz with a 4th order LWR low pass. Mains play down to 50Hz and you apply a 2nd order LRW high pass at 30Hz to protect them from any sub bass. Then you can apply shelving that centres around 75Hz and lowers the output of the sub from 50-100Hz (reaching maximum shelf down at 100Hz). Then apply a similar, but opposite, shelf on the mains at 75Hz where they reach the maximum shelf down by 50Hz. Or you can use notches.
The numbers I've chosen arbitrarily just to highlight what I'm talking about.
If you want to do anything to improve the quality of the bass in your room then what you should really be doing is adding another sub and not changing the one you've already got. As you've already got one (the ultimax 2) that can cover down to 20Hz you do not need another. You only need secondary subs that have bass extension down to as low as your rooms lowest room mode. This usually happens around 30-40Hz. So go build a 10-12" sub and run it together with the UM2 up to 100Hz.
They have a pickle of a resonance at about 330Hz and this will probably trickle down into increased third order distortion at about 110Hz but provided you've used a 4th order targeted at 100Hz this should be fine.
The main thing about bass reproduction is that it's dominated by the room and much less so by the actual speakers themselves. At least so long as the speakers (subs) aren't poorly designed.
When you're altering the crossover frequency, between the sub and your mains, you're also altering the interplay of the room modes from the various bass sources and their positions within the room. More bass sources is better. More bass sources that are spread into different locations are even better.
The higher you take the single sub the worse, and more defined, the room modes are going to be over the region that the sub is covering. That being Vs the two separate bass sources of your mains.
Changing the sub isn't going to solve this. Ideally you'd want to run the sub up to 100Hz and the mains down as low as they'll happily go. This creates an overlap between them and provides you with 3 separate bass sources. Obviously this will create an excess of output where the 3 overlap but this is simple enough to fix using some shelving filters or PEQ.
As an example sub plays up to 100Hz with a 4th order LWR low pass. Mains play down to 50Hz and you apply a 2nd order LRW high pass at 30Hz to protect them from any sub bass. Then you can apply shelving that centres around 75Hz and lowers the output of the sub from 50-100Hz (reaching maximum shelf down at 100Hz). Then apply a similar, but opposite, shelf on the mains at 75Hz where they reach the maximum shelf down by 50Hz. Or you can use notches.
The numbers I've chosen arbitrarily just to highlight what I'm talking about.
If you want to do anything to improve the quality of the bass in your room then what you should really be doing is adding another sub and not changing the one you've already got. As you've already got one (the ultimax 2) that can cover down to 20Hz you do not need another. You only need secondary subs that have bass extension down to as low as your rooms lowest room mode. This usually happens around 30-40Hz. So go build a 10-12" sub and run it together with the UM2 up to 100Hz.
Trust me ive been a DJ for 15 years, and I know my room and system quite well, including the Daytons. I know what I want from sound. They are now doing well, good actually. Well integrated. Ive tested for weeks, months different configs and placements, rooms treatments, DSP, MSO, DIRAC, and what not. Im very satisfied, especially with the help from some great masters of subwoofer applications. Now its time for something different. Even if I want some of the Daytons low back, I wont have or need their big 15" woofers and cabinets with it.
The only way to get really good bass, in a domestic sized environment, is with multiple bass sources correctly integrated and equalised. You'll never, and I repeat never, get there with anything less.
If none of these subwoofer masters have correctly lead you down the multiple sub trajectory, a la Earl Geddes, then they haven't been particularly good masters.
If none of these subwoofer masters have correctly lead you down the multiple sub trajectory, a la Earl Geddes, then they haven't been particularly good masters.
I've never used subbox.pro, so I think that you'll need to experiment with a design or two in order to ascertain what it can do.
The frequency response results for the 18SW115-8Ω and 18SW115-4Ω in the 160-litre enclosure tuned to 28Hz are very close to each other in the simulations. I think that the main reason to choose the 18SW115-4Ω model is that you will have more power in reserve for reproducing short term transients.
The iNUKE NU6000DSP amplifier is capable of 2x3100W into 4Ω, which seems like more than enough power on tap. It is more than capable of getting the driver to exceed its Xmax at low frequencies.
The manual states:
- The PEQ module deploys up to eight EQ bands for precise sound sculpting. The EQ bands can each be switched between low shelving, high shelving, and parametric modes. For the high shelving and low shelving EQ bands, the LS12 and HS12 settings provide steeper equalization curves than the LS6 andHS6 settings.
- The XOVER module offers programmable pairs of high- and low-pass filters. ... Each low-pass and high-pass filter also offers multiple options for filter type and slope: OFF, Butterworth (BUT6, BUT12, BUT18, BUT24, BUT48), Bessel (BES12, BES24), or Linkwitz-Riley (LR12, LR24, LR48).
- The BEHRINGER Amp Remote software allows the user to control all iNUKE DSPsettings remotely from a computer via the USB connection located on the iNUKEfront panel.
By applying an appropriate low-pass shelving filter, we can get the following response. Notice how flat it is. I was a bit surprised how well it worked. The response is −3dB at about 26Hz or so.
We get a very nice low-pass filtered response when we add a 4th-order Linkwitz–Riley low-pass filter set to 100Hz. This subwoofer should blend in very cleanly and smoothly with the main speakers when they use a complementary 4th-order Linkwitz–Riley high-pass filter.
@witwald
I would caution you about Behringer's power ratings for the NU/NX series amps. They are pure BS. Here is a link to measurements of the NU6000:
https://www.avsforum.com/threads/measuring-amplifiers.1780914/page-27?post_id=33179737#post-33179737
Read through the thread for complete details. Some example measurement results:
at 10Hz --> 1275W
at 20Hz --> 1375W
at 30Hz --> 1475W
at 40Hz --> 1530W
at 100Hz --> 1890W
at 1kHz --> 2500W
This is power per channel, two channels driven, with a load of 4 Ohms on each channel. The measurement signal was a "burst", meaning very brief in duration (only a couple of cycles).
For subwoofer duty (below 100Hz) these values are well below 3.1k W per channel that you (and Behringer) claim. Also, when the amp must delivery this much power for several seconds the power level will fall further by IMO about 50%. Near-continuous low frequency signals are rare except for the most low crest factor bass heavy music, but it is good to keep this in mind. The power supply of the amp is just not capable of meeting the demands of high power at low frequencies.
I own a couple of NU3000 with quiet fans and I use them for sub duty sometimes. The NU and NX amps from Behringer are a great value, but you have to accept that they are not capable of the MFG power ratings.
I would caution you about Behringer's power ratings for the NU/NX series amps. They are pure BS. Here is a link to measurements of the NU6000:
https://www.avsforum.com/threads/measuring-amplifiers.1780914/page-27?post_id=33179737#post-33179737
Read through the thread for complete details. Some example measurement results:
at 10Hz --> 1275W
at 20Hz --> 1375W
at 30Hz --> 1475W
at 40Hz --> 1530W
at 100Hz --> 1890W
at 1kHz --> 2500W
This is power per channel, two channels driven, with a load of 4 Ohms on each channel. The measurement signal was a "burst", meaning very brief in duration (only a couple of cycles).
For subwoofer duty (below 100Hz) these values are well below 3.1k W per channel that you (and Behringer) claim. Also, when the amp must delivery this much power for several seconds the power level will fall further by IMO about 50%. Near-continuous low frequency signals are rare except for the most low crest factor bass heavy music, but it is good to keep this in mind. The power supply of the amp is just not capable of meeting the demands of high power at low frequencies.
I own a couple of NU3000 with quiet fans and I use them for sub duty sometimes. The NU and NX amps from Behringer are a great value, but you have to accept that they are not capable of the MFG power ratings.
Thanks guys. I totally subscribe take anything behringer states with a grain of salt. Some guys said to drive this 18SW115 you need 2 6000inuke DSP both in bridged mode each driving one sub.
Since they wont see any 100+ people party but strictly home use I think the one Behringer inuke 6000 will do. Im actually just waiting for it to fail - many people complain about reliability, but since I dont push it so hard it still works well after 6 months. I also have one Crown XSL 1502 but that is way underpowered for this, maybe only in bridged mode. So ill use the 6000.
I wont use its DSP, that I do with MiniDSP 2x4HD and use MSO created filters and XO. This gives good and easy calcuclations without phasing or time issues.
I did manage to create something in subbox.pro. I chose a 34Hz tuning freq since I couldnt get your 28hz suggested tuning frequency to work well over the entire range to 100hz. Can you run a 34hz config through your software @witwald and either confirm or warn me not to do that and instead go for the 28hz version? I feel its getting close to the drivers limits and close to a high pass filter I need to implement.
This design I ade in subbox.pro. However, I dont see any bracings. Does it matter for the final design much if you add bracings? I guess in a ported design you dont add wool or only strictly to the walls?
A last point: B&C also has a 18SW100, wich is a lower specced brother of the sw115 but the same family. Since these sw115 go so over any spl ill ever need (i will have a 105 max config and only with earplugs in my house for short moments of techno or rock or whatever). Would it be wise to look at that SW100 instead even if it has a slightly higher Fs, or stick with this sw115 since both already have such a higher low limit than my current daytons. Its actually also a financial consideration. This is more of an experiment and the sw100 are almost 200 euro less than the sw115's.
Since they wont see any 100+ people party but strictly home use I think the one Behringer inuke 6000 will do. Im actually just waiting for it to fail - many people complain about reliability, but since I dont push it so hard it still works well after 6 months. I also have one Crown XSL 1502 but that is way underpowered for this, maybe only in bridged mode. So ill use the 6000.
I wont use its DSP, that I do with MiniDSP 2x4HD and use MSO created filters and XO. This gives good and easy calcuclations without phasing or time issues.
I did manage to create something in subbox.pro. I chose a 34Hz tuning freq since I couldnt get your 28hz suggested tuning frequency to work well over the entire range to 100hz. Can you run a 34hz config through your software @witwald and either confirm or warn me not to do that and instead go for the 28hz version? I feel its getting close to the drivers limits and close to a high pass filter I need to implement.
This design I ade in subbox.pro. However, I dont see any bracings. Does it matter for the final design much if you add bracings? I guess in a ported design you dont add wool or only strictly to the walls?
A last point: B&C also has a 18SW100, wich is a lower specced brother of the sw115 but the same family. Since these sw115 go so over any spl ill ever need (i will have a 105 max config and only with earplugs in my house for short moments of techno or rock or whatever). Would it be wise to look at that SW100 instead even if it has a slightly higher Fs, or stick with this sw115 since both already have such a higher low limit than my current daytons. Its actually also a financial consideration. This is more of an experiment and the sw100 are almost 200 euro less than the sw115's.
@elmoggbt7736633 I am afraid that "some guys" are wrong about these amps. The NU6000 or NX6000 amps (DSP or not) cannot be bridged. There are two channels. Each "channel" of the 6000 series is essentially a bridged NX3000 amplifier, so cannot again bridge those channels. The power measurements I linked to is all you can get out of it.
This is the b&c 18sw100 drawing. It needs more liters, on a 31hz tuning freq. it looks very simple and easyto make, even for a guy like me (im disabled, in a wheelchair. However, i got the dayton diy parts express kit glued together in 20 minutes, working fine.)
How about this @witwald : I got ChatGPT a bit tighter on the leash and it made a comparisment between your 160liter/28hz tune suggestion and my WInISD conclusion. Turns out indeed the 28hz is a slightly but nevertheless better design.
However, chatgpt cant figure out a port for the 28hz config unless the box becomes too big. It tested air velocity and says a 65x65x x depth box still makes way too much air velocity
However, chatgpt cant figure out a port for the 28hz config unless the box becomes too big. It tested air velocity and says a 65x65x x depth box still makes way too much air velocity
Last edited:
And its finished, and double checked. Were going to build a 28hz tuned 160 liter box for the b&c 18SW115 on @witwald design calculation this is the best it can do in this config.
The front vent is 011x0.4m with total lenght of 87cm, so folded. In a 64x60x60 box. Winisd calculated about 19ms air velocity at 1000 watt, about 22 at 1400 and a max of 24.5 at full 1700 watt power. Is this responsible design enough? The design is from subbox, chatgpt did the calculations, and later made some bracings suggestions.
The front vent is 011x0.4m with total lenght of 87cm, so folded. In a 64x60x60 box. Winisd calculated about 19ms air velocity at 1000 watt, about 22 at 1400 and a max of 24.5 at full 1700 watt power. Is this responsible design enough? The design is from subbox, chatgpt did the calculations, and later made some bracings suggestions.
Last edited:
I would try and avoid using a right-angled L-bend for the port. It's likely to generate some turbulence at the corner.
In any case, the port length Lv = 47.2 cm calculated by VituixCAD is much shorter than the Lv = 79.7 cm that your box drawing shows. Maybe we are using different end corrections to compute the length of the slot?
In any case, the port length Lv = 47.2 cm calculated by VituixCAD is much shorter than the Lv = 79.7 cm that your box drawing shows. Maybe we are using different end corrections to compute the length of the slot?
Indeed but bear with me, I didnt even know what WinISD was until last week and only got grasb of it since the last 2 or 3 days.
I examined your images carefully and noted instantly you have a different end correction. I tried to insert that in WinISD but it keeps jumping back to 0.732. After your comment I went and tried again, googled some more and it seems the current version of WinISD doesnt let people change either port lenght or end correction. Why I dont know but there should be a big warning as this has huge implications for anyone trying to build a vented design.
The advice is to work further in Hornresp, another program I only discovered a few days ago.
I also looked at some other designs like Keystone or Cubo but reflex remains best compromise for home and music use.
Thanks for the help and corrections. To be continued!
I examined your images carefully and noted instantly you have a different end correction. I tried to insert that in WinISD but it keeps jumping back to 0.732. After your comment I went and tried again, googled some more and it seems the current version of WinISD doesnt let people change either port lenght or end correction. Why I dont know but there should be a big warning as this has huge implications for anyone trying to build a vented design.
The advice is to work further in Hornresp, another program I only discovered a few days ago.
I also looked at some other designs like Keystone or Cubo but reflex remains best compromise for home and music use.
Thanks for the help and corrections. To be continued!
Actually, subbox.pro has the same 0.73 end correction. It keeps making a 79cm lenght port. Any suggestions where el;se i can make a design but with Vituixcad port lenght and 2.2 end correction?
0.73 end correction is for a round port that is close to center locations of a box. The 2.2 is for slot port corrections.
Evenso, winISD doesnt let me change. It does let you change to vent or slot, but the end correction doenst change. But Subbox pro also doesnt allow such. Which leaves me confused. Ive tried sketchy or speakerbox lite but enter lots of problems and complications.
Cant I just cut the drawing above, the 53cm board to 47 and leave out the 14cm folded part? Or do I need adjustments elsewhere to stay at Witwalds 440cm2?
Cant I just cut the drawing above, the 53cm board to 47 and leave out the 14cm folded part? Or do I need adjustments elsewhere to stay at Witwalds 440cm2?
It's possible to use an indirect approach to get SubBox to change the length of the slotted vent. All you need to do is adjust the vent tuning frequency in SubBox until the length of the slot matches the one that was computed by VituixCAD. This will need to be done iteratively, by trial and error. A few iterations should enable you to get quite close to the desired slot length.
Yes good suggestion. I was thinking, maybe I dont need 440cm2 port area and can make do with a bit smaller port area too. The calculations have the air velocity around 19ms at 1000watt power but the speaker will be doing aprox 127db and I dont think my marriage will survive playing at those levels.
What if i consider a max of 600 watt, and make it 340cm2 port area. According to my calculations, it would still need to reach about 680 watt and still putting out close to aprox 119db before any chuffing would appear. Can I take that gamble and consider I made a box that is not optimal above 680watts? Or is that bad engineering? Im just playing around and figure if I can get the opening a bit smaller.
What if i consider a max of 600 watt, and make it 340cm2 port area. According to my calculations, it would still need to reach about 680 watt and still putting out close to aprox 119db before any chuffing would appear. Can I take that gamble and consider I made a box that is not optimal above 680watts? Or is that bad engineering? Im just playing around and figure if I can get the opening a bit smaller.
Edit: nevermind. Im starting to have 2nd thoughts, mainly about the switch to ported designs. Maybe I can investigate the GSG suggestion: keep the daytons for low, and add a more mid bass strong motor oriented b&c for mid bass in a sealed config. The DS115 seems way overkill and so im looking at the 15DS100. What do you think Witwald? Does that make sense? Could you run it through your program and see if and if yes, where we could marry these drivers? Let the 15" Ultimax run from 20-50 and from there have the 15DS100 cover till 110 or something in a nice small box of say 55 liters. I have minidsp, I have mso. I have a lot of amp power. The mains have some placement - room issues betwen 110-150 and so maybe I could even make that DS just take over to about 150. DSP all good to tame that DS motor. Either way, Im gonna put brakes on plans for now and do some more reading, think about alternatives until im 100% sure.
Last edited:
You've mentioned that the "Daytons seem not to be as tight above 70Hz", and "so far the best sounding is around 100Hz". It's not entirely clear how one would assess the "tightness" of the response above 70Hz. You may be experiencing a subtle, but nevertheless present, dip in the response in the 70Hz region.
This might be caused by a mismatch between the low-passed and high-passed driver responses. Even a small difference of 0.5dB will reduce the energy that is being dispersed by 12%. I can well imagine that such a reduction will come across as a significant change in "tightness".
To investigate this issue, you could simply add a bit of parametric EQ to the response to get back some of the lost tightness. Simply add +0.5dB at 70Hz with a Q=2.0. Does this give you back some of the missing "tightness"?