need advice on box for a xls12 sub

[shannow]

Hello i'm french and I need some advice for a box for a xls12 Peerless . I don't know if you have the specs but here there are :

SWR 308
Thiele Small parameters: Free air Common Baffled
Nominal impedance Zn (ohm) 8
Minimum impedance/at freq. Zmin (ohm/Hz) 4.7/106
Maximum impedance Zo (ohm) 64.2
DC resistance Re (ohm) 3.5
Voice coil inductance Le (mH) 4.2
Capacitor in series with 8 ohm
(for impedance compensation) Cc (µF) 38
Resonance Frequency fs (Hz) 18.1 18.1
Mechanical Q factor Qms 3.70 3.70
Electrical Q factor Qes 0.21 0.21
Total Q factor Qts 0.20 0.20
F (Ratio fs/Qts) F (Hz) 90
Mechanical resistance Rms (Kg/s) 5.12
Moving mass Mms (g) 166.3 166.4
Suspension compliance Cms (mm/N) 0.46
Effective cone diameter D (cm) 24.4
Effective piston area Sd (cm²) 466
Equivalent volume VAS (ltrs) 139.2
Force factor Bl (N/A) 17.6
Reference voltage sensitivity
Re 2.83V 1m at 106 Hz (Measured) (dB) 90.6

Magnet and voice coil parameters:
Voice coil diameter d (mm) 51
Voice coil length h (mm) 33
Voice coil layers n 4
Flux density in gap B (T) 1.04
Total useful flux (mWb) 2.50
Height of the gap hg (mm) 8
Diameter of magnet dm (mm) 147
Height of magnet hm (mm) 35
Weight of magnet (kg) 2.42

Anyway I want to make the sub to complete my celius triangle 202 who needs subs ... (it's conected with a rotel ra-01) . And so the main use for this system would be 80% for music (metal ;hard rock ; orchestra ,newage) and 20% for games .

And the question is: do you think that, for mainly audio purpose , a 40L vented with a 3inch port (9inch long) ,tuned at 35hz , would do the job ?
i'm looking for the drum , bass ,punch and the shacking the room sensation when a symphonic orchestra is played with deep bass.
Many thanks in advance .

 

[paulspencer]

I ran some simulations through WinISD professional for some different alignments.

All of them have a rumble filter @ 20 Hz and a low pass at 80 Hz 2nd order

Sealed 30L with Linkwitz transform - blue
50L with 2x passive radiators - green
50L with 25 Hz vented fb - red
40L tuned at 35Hz - white

All shown with an input power that will not exceed excursion above 20 Hz.

Notice how you can get more extension than with your 40L box which is tuned a little high for a sub? Given the higher tuning you can get some more output around 40 Hz, but you lose some depth.

The sealed sub loses about 8db if but has the same extension (in fact slightly more). At 40 Hz you only lose about 3db output with the sealed box.

Watch that you flare both ends of your vents or you will have turbulence issues.

 

[richie00boy]

A 3 inch port is too small for an XLS12, especially tuning as high as that.

 

[paulspencer]

A 3 inch port is too small for an XLS12

With generous flares it will probably be sufficient, although bigger would be an advantage if it can be accomodated.

especially tuning as high as that.

Other way around if I'm not mistaken - lower tuning means more air in the vent, hence more turbulence and a bigger vent required. This is why a rumble filter helps reduce turbulence which is caused by very large cone excursions with sub bass.

 

[richie00boy]

Nope. Small published a method of working out the minimum vent surface area, and the area required goes down with frequency.

 

[THOR]

critical Q sub

here try this. highly recomended🙂

http://members.ozemail.com.au/~joeras/sub_index.htm

 

[paulspencer]

Nope. Small published a method of working out the minimum vent surface area, and the area required goes down with frequency.

What's the name of the paper and where can it be obtained?

 

[paulspencer]

The sub that Thor linked uses this driver and shows measurements lower in distortion than a velodyne servo controlled sub, which is quite impressive. Linkwitz also has a similar design on his site. This is quite an endorsement of this driver in a sealed box. This driver isn't often used in a vented alignment due to its small size, hence the difficulty getting a decent vent into it.

 

[paulspencer]

Nope. Small published a method of working out the minimum vent surface area, and the area required goes down with frequency.

I find this hard to believe. I don't wish to question Small, but since I don't have his paper right in front of me ...

As you go down in frequency more air is being displaced, therefore vents need to be bigger. The worst turbulence is caused by large excursions at sub bass frequencies.

I did a quick little test. This may not be the final word on the matter, but gives an indication ...

40L vented box with XLS and 3" vent

tuned @ 35 Hz ... vent mach 0.15
tuned @ 30 Hz ... vent mach 0.18
tuned @ 25 Hz ... vent mach 0.21

According to WinISD the box tuned higher has acceptable vent air speed, but the latter 2 do not.

Better than any of these would be a 50L box tuned @ 25 Hz with a 4" vent with a vent mach of 0.11

 

[richie00boy]

I find this hard to believe. I don't wish to question Small, but since I don't have his paper right in front of me ...

The paper is 'THE' paper written by Small following Thiele's one on vented boxes. It will be available soon (days I hope) on my webspace and I will publish a link to it.

As you go down in frequency more air is being displaced, therefore vents need to be bigger. The worst turbulence is caused by large excursions at sub bass frequencies.

Your thinking is not entirely correct. You are confusing excursion required to supply same SPL as frequency goes lower. In your case, the excursion increases with frequency lowering.

However, for a given excursion level, a reduction in frequency correlates to a reduction in the vent area required.

 

[paulspencer]

You are confusing excursion required to supply same SPL as frequency goes lower. In your case, the excursion increases with frequency lowering.

However, for a given excursion level, a reduction in frequency correlates to a reduction in the vent area required.

I don't believe I am confusing anything, although I await to see this link to the article ...

 

[richie00boy]

I know what you are thinking about going lower requires more excursion so more port movement -- that's true, but it's also a red herring in this particular case. You need to think in terms of a single given excursion level to understand how it works.

OK, here's another bit of weight to my argument, in lieu of the availablility of the paper. As frequency goes down, the velocity of the air in the port reduces. Because the airspeed is reduced the chances of it creating noise are also reduced.

 

[shannow]

I'm very happy to see so many repplies and very usefull thanks 😉 So a 3inch vent is too small ... I've just been playing around with winisd and I can't see how a so big vent can fit a 40litre small box for this speaker :s . I'm still learning how to use winisd but i've found a pretty nice frequency response with a 70L tuned at 25hz , a 10cm diameter vent and 50cm long (i've used a lowpass butterworth at 80hz...) Is it good ? thanks 😉

 

[paulspencer]

OK, here's another bit of weight to my argument, in lieu of the availablility of the paper. As frequency goes down, the velocity of the air in the port reduces. Because the airspeed is reduced the chances of it creating noise are also reduced.

Starting to make more sense ... lower frequency means more cone movement but at the same time higher frequency means more cycles - two things happening at the same time which would have to be considered together. But still if this is so then WinISD got it wrong ...

Again, I'm quite interested to see this paper

 

[paulspencer]

70L sounds a bit big. I've never heard of anyone using the XLS in such a big box. 28 - 50L is typical for this driver as it's designed to work in a small box, either sealed or passive radiator. Vented is a bit more tricky.

Perhaps see how you go with 3" or 4" in a 50L box, with both ends generously flared.

 

[richie00boy]

Starting to make more sense ... lower frequency means more cone movement

That is true, but as I tried to point out previously it's a red herring. You must forget about increasing excursion with reducing frequency and be concerned only with the maximum excursion you intend to use. Here's an illustration off the top of my head with guessed -- but indicative -- numbers:

Running the driver at 5mm excursion *in each case* (that's the important bit)

20Hz tuning: Min Sv = 100 sq cm (lowest vent velocity)
25Hz tuning: Min Sv = 120 sq cm
30Hz tuning: Min Sv = 140 sq cm (highest vent velocity)

Yes, going down in frequency the excursion will increase, but ONLY TO MAINTAIN THE SAME SPL. When sizing a vent, we don't care about maintaining SPL across a bandwidth; we are trying to determine the WORST CASE which will be the point of MAXIMUM EXCURSION.

In your modelling you are failing to make the excursion the same in each case, you're just modelling differently tuned boxes fed with the same signal. Of course, then you find that excursion increases as you go down in frequency, BUT IS NOT THE SAME EXCURSION FOR EACH TUNING FREQUENCY, which has (what I now hope should be obvious) effect on the vent requirement. You must adjust the signal to achieve the same excursion at the point of interest.

🙂 😎

 

[shannow]

Yes you're right 70L might be a bit big for it as most people don't go over 50L for this driver 🙁 .... what i'm trying to do is to imitate the shiva's adire frequency response with a sealed 88.5L(Q0.6) ,and a lowpass filter 80hz ,wich I know is very good for audio purpose . The closest I've gotten so far ( winisd )with the XLS12 is with a 4th order bandpass i'm starting getting very close with no eq/filters . AT the moment i'm trying with a base like this(for the 4th order pass band) : rear chamber (sealed) 50L(35.5hz) , and the front chamber 15L with tuning freq at 55hz . I'll try to improve it again 😉 .

 

[richie00boy]

shannow,

I'm afraid that if you are trying to imitate that frequency response then you are wasting your time. The frequency response is important, but with it also comes a particular phase and transient response and group delay. A frequency response for a sealed box will have a totally different phase and transient response and group delay to a vented box with the same (OK, it will never be the same really, but just considering say -3dB bandwidth) frequency response. The only way you will make an XLS sound like that Shiva is with a sealed box and Linkwitz Transform circuit to achieve the same Qtc and F3.

Because the XLS has a low Qts and fairly low Vas, the box will be small. You will have a massive problem (if not impossible) to accomodate the vent necessary to for good performance. This drive unit is designed to be used with a passive radiator, or at a pinch sealed with EQ.

If you make a large vented box then the transient performance will suffer and power handling potential (loudness) not reached.

 

[paulspencer]

That doesn't sound like an ideal benchmark to me. Instead I would suggest something that would get an F3 with room gain of about 25 Hz. You could expect in a typical room a gain below 40 Hz of approx 6db/octave. So a typical sealed box yields:

F3 @ 40 Hz

anechoic
40 Hz : - 3 db
20 Hz : -15 db

in room
40 Hz : - 3 db
20 Hz : - 9 db

For music you might not consider this necessary at all. In fact you may prefer it without this eq.

The 50L - 25Hz tuned version appears to work quite well with the room gain mentioned above. It achieves an in room F3@ 23Hz.

I would stay away from bandpass for music. Either a small sealed box with eq if desired or a vented/passive radiator box a little bigger. 70L for this driver is like an EBS!

A 90mm vent in a box this size would need to be 530mm long, with generous flares on both ends, and would only need one bend.
4" >>>> 670mm long
90mm vent in 40L box >>>> 690mm long
100mm vent in 40L box >>> 890mm long

As you can see, the small box makes it very difficult to get a vent in. Using a box slightly larger than you would otherwise require and a 90mm vent makes it easier.

If you can get the vent to work well regarding turbulence etc then it should have comparable performance to a passive radiator. I'd experiment with this before getting a passive radiator (which models slightly deeper extension with 2 of them) as you may like the sound of sealed better for music.

 

[shannow]

yeah it's a pity then ....well I haven't bought the speaker yet I was very interrested with a shiva adire (88.5L sealed ) but It's very difficult to get hold of one in europe 🙁 I've been looking at cars speakers .... at the moment none have matched what the shiva can do 🙁 .You live in uk richie00boy ?

what speaker would you advice me for about 100£?

I'm going to england in a week (south east area and perhaps near london ...) do you know a good shop in that area?
thank you again for all your advices

🙂