How to deal with high QTC?

[soren5]

I need your help.

I have two 10 liter (0.35 cu ft) cabinets and two Peerless SLS 8" woofers and would like to make some low-end stereo support for my sealed Alpair 7 full-range speakers - crossed at ~300Hz. This is for music only in a 6x6x3 meter (19x19x9 ft) room. Power handling of the woofers is quite good and I got powerful amps, so LT is also a possibility. I'm going active and using a miniDSP for the crossover.

The problem is that WinISD gives me a QTC of 1.1 with this driver in a 10 liter box (probably QTC will be a bit lower when filling, but still far from 0.707). So how should I go about this in the best way?

I see three possibilities:

1) The EAS/ELF/Bagend approach where you utilize the high QTC and add EQ. I'm not sure how this will work (if it will work at all) with a 300Hz XO.

2) The aperiodic enclosure approach. I could add glass wool stuffed “vents” to the cabinets and in this way lower the Q.

3) A combination of 1 and 2.

What are your thoughts on this? I've attached the driver specs and also a WinISD frequency response of the driver in a 10 liter box.

Advice appreciated

 

[planet10]

I would try 2

dave

 

[soren5]

I knew you were going to suggest that, Dave

But does the aperiodic enclosure prevent me from boosting the low end? If going aperiodic could lower the QTC to 0.707 I would get an f3 of 48Hz, which is still not amazingly deep.

The woofer's relatively high power handling and excursion tells me I could improve the low end by adding a LT.

 

[chris661]

I'd be wary of using a LT. WinISD has the capability of adding one, but be sure to check the apparent amplifier load - this gives an indication of how much power they'll be taking right at the bottom end. Excursion gets scary, too.
That said, a Linkwitz Transform offers (I suspect) more flexibility to pick and choose Qtc & LF extension than an aperiodic design.

Chris

 

[planet10]

But does the aperiodic enclosure prevent me from boosting the low end? If going aperiodic could lower the QTC to 0.707 I would get an f3 of 48Hz, which is still not amazingly deep.

But what is F10? It is a small driver in a small box, you can't expect to go too low.

dave

 

[weltersys]

I need your help.

I have two 10 liter (0.35 cu ft) cabinets and two Peerless SLS 8" woofers and would like to make some low-end stereo support for my sealed Alpair 7 full-range speakers - crossed at ~300Hz. This is for music only in a 6x6x3 meter (19x19x9 ft) room. Power handling of the woofers is quite good and I got powerful amps, so LT is also a possibility. I'm going active and using a miniDSP for the crossover.

The problem is that WinISD gives me a QTC of 1.1 with this driver in a 10 liter box (probably QTC will be a bit lower when filling, but still far from 0.707). So how should I go about this in the best way?

I see three possibilities:

1) The EAS/ELF/Bagend approach where you utilize the high QTC and add EQ. I'm not sure how this will work (if it will work at all) with a 300Hz XO.

2) The aperiodic enclosure approach. I could add glass wool stuffed “vents” to the cabinets and in this way lower the Q.

3) A combination of 1 and 2.

What are your thoughts on this? I've attached the driver specs and also a WinISD frequency response of the driver in a 10 liter box.

Advice appreciated

The 10 liter box shows a bumped up response above F3, what does the QTC look like in a larger box?

 

[cyclecamper]

So you need a bigger box. My favorite solution to that problem:

Get two more of the same woofers, and mount each pair clamshelled (drivers facing each other) isobaric-style in a 10-liter box. That gives you a "Q" like a box twice the size, and a much shallower rolloff. Spend on drivers, save on wood?

Then you can add EQ until the xmax limitation is reached, and it will handle (and need) more power.

But that assumes you don't mind the sight of a magnet sticking out into the room.

 

[weltersys]

So you need a bigger box. My favorite solution to that problem:

Get two more of the same woofers, and mount each pair clamshelled (drivers facing each other) isobaric-style in a 10-liter box. That gives you a "Q" like a box twice the size, and a much shallower rolloff. Spend on drivers, save on wood?

Then you can add EQ until the xmax limitation is reached, and it will handle (and need) more power.

But that assumes you don't mind the sight of a magnet sticking out into the room.

And don't mind the chuffing noises coming from the vent hole...

 

[sreten]

Hi,

Maximum SPL in the excursion region of a sealed driver only depends on cone
area and excursion, not box size or response. If you curtial the response
then you need more power to reach the excursion limits, if you EQ the
response for deeper bass (an LT) this needs more power and overall
SPL (due to the extended bass) limits will be lower.

Aperiodic loading is not a good idea, it chucks away cone excursion by
allowing vent output to cancel driver output, its not wanted for a sub,
and no sub I'm aware of uses it for this reason.

I can't see anything wrong with the LT approach, in moderation, LT's
are all about good bass extension but at inevitably limited SPL levels.

Limited SPL ? did somebody mention FR drivers ?

rgds, sreten.

 

[Pallas]

And don't mind the chuffing noises coming from the vent hole...

The SLS 8 is pretty quiet. And it doesn't have a standard pole-vent hole. Pole venting is through a radial holes under the dustcap.

Still, it's a pretty deep woofer. So deep that one may as well just make the box wider, and thus bigger.

 

[revboden]

I think if you get to 35Hz F6 with a 8" driver you're doing really well. People expect too much of Xmax in a small driver.

Not really directed at any comment, just sayin

 

[planet10]

Aperiodic loading is not a good idea, it chucks away cone excursion by
allowing vent output to cancel driver output, its not wanted for a sub,
and no sub I'm aware of uses it for this reason.

You only need 1 counter-example to disprove something... i've built a couple VERY good subs that are aperiodic. Both push-push 12" in a box that, if sealed, would have pushed the Q higher than i wanted.

dave

 

[head_unit]

...probably QTC will be a bit lower when filling, but still far from 0.707

As a speaker engineer, let me say you *may* be wrong. You can theoretically get like 40% more volume. For actuals, Vance Dickason's Loudspeaker Cookbook has a BUNCH of tests (and tons of other info), so order that from AudioXpress.com.

Pick the best stuffing and just TRY IT. Parameters are not very exact and neither are most simulators. You have the woofers and you have the cabinets, so nothing to lose just putting it together and listening.

If still not satisfied, then sorry, you will simply need bigger cabinets. In my long experience, there are funny tricks (complex passive filters included) but no real working shortcuts to a properly sized enclosure. That means forget the aperiodic thing, and you are way away from the Bag End thing. That would be with a resonance well above 100 Hz and a crossover below that, not 300 Hz. The Bag End deal is the sub is operating ENTIRELY down on the rolloff slope --> compact but hugely power hungry.

 

[GM]

The problem is that WinISD gives me a QTC of 1.1 with this driver in a 10 liter box (probably QTC will be a bit lower when filling, but still far from 0.707). So how should I go about this in the best way?

Hmm, net Vb will be around 9 L and comparing it damped to 0.707 Qtc in WinISD Pro to MJK’s MathCad software using polyfil, it looks like stuffing it with ~ 0.317832 lbs [~144.2 g] will do, though its high inductance will still give it a bit of a ‘bump’ in its response around 100 Hz unless doubled [~188.4 g].

Of course, all this assumes the published specs are accurate and the box is truly leak free, so strictly a starting point.

GM

 

[sreten]

You only need 1 counter-example to disprove something...
i've built a couple VERY good subs that are aperiodic.
Both push-push 12" in a box that, if sealed, would
have pushed the Q higher than i wanted.

dave

Hi,

I of course meant a commercial design. Of course you can build an
aperiodic sub if you want, but you are chucking away excursion.

Trying to go from Qbox = 1.1 to Qbox = 0.7 aperiodically
won't work as well (if at all, its a big change), an LT will.

rgds, sreten.

 

[tb46]

The answer (if the link works) is:

http://www.everwonder.com/david/tacobell/videos3/bquiero8b.avi

Regards,

 

[soren5]

Thanks for all the comments. I had never imagined my thread would cause so much activity! But it seems that the answer is not so simple after all. There are for sure different opinions about what is the best approach (not considering the obvious best approach, which would be to get a larger box).

One thing that seems to me clear though is that the Bagend approach won't work if I'm crossing at 300Hz.

I'll try to answer or comment on some of your posts.

The 10 liter box shows a bumped up response above F3, what does the QTC look like in a larger box?

You mean what the frequency response looks like when QTC=0.707? WinISD says this QTC would require a 52 liter box and I've attached the graph below.

But what is F10? It is a small driver in a small box, you can't expect to go too low.

F10 is ~28Hz for QTC=0.707 and for QTC=1.1 (10 liter box) I get F10 ~42Hz.

I think if you get to 35Hz F6 with a 8" driver you're doing really well. People expect too much of Xmax in a small driver.

F6 is ~37Hz for QTC=0.707

As a speaker engineer, let me say you *may* be wrong. You can theoretically get like 40% more volume. [...] Pick the best stuffing and just TRY IT. Parameters are not very exact and neither are most simulators. You have the woofers and you have the cabinets, so nothing to lose just putting it together and listening.

This approach makes sense to me But still there is a long way from 10 to 52 liters! (A quick calculation tells me this is more than 40% ) I'll try with stuffing and see how it works, and if not, I guess I have to modify the boxes. What would you recommend as the "best" stuffing here?

Hmm, net Vb will be around 9 L and comparing it damped to 0.707 Qtc in WinISD Pro to MJK’s MathCad software using polyfill, it looks like stuffing it with ~ 0.317832 lbs [~144.2 g] will do, though its high inductance will still give it a bit of a ‘bump’ in its response around 100 Hz unless doubled [~188.4 g].

So you're saying that provided the box is well sealed and the driver specs are correct, I could get a flattish response with 188g polyfill in a 10 liter box?

Trying to go from Qbox = 1.1 to Qbox = 0.7 aperiodically won't work as well (if at all, its a big change), an LT will.

Does this mean that a LT can both "remove" a possible 100Hz bump as well as boosting the lower end?

 

[weltersys]

Thanks for all the comments. I had never imagined my thread would cause so much activity! But it seems that the answer is not so simple after all. There are for sure different opinions about what is the best approach (not considering the obvious best approach, which would be to get a larger box).

One thing that seems to me clear though is that the Bagend approach won't work if I'm crossing at 300Hz.

Does this mean that a LT can both "remove" a possible 100Hz bump as well as boosting the lower end?

LT is simply a corrective EQ designed for a simulated response. Why not simply use a graphic or parametric EQ and adjust for the actual in room response of your speaker?

The Bagend approach uses corrective EQ with limiting to keep the huge LF boosts needed from exceeding Xmax, which works OK at low levels.

The 52 liter box has more than 6 dB level at 30 Hz than the too-small box, it would require only 1/4 the power to correct the response down low.
To put it another way, the larger box will require the same power to EQ flat to 30 Hz as the small box to about 45 Hz.

There is plenty of musical content below 45 Hz, the low B on a five string bass is about 31 Hz. It takes about 104 dB to hear 30 Hz at the same perceived level as 80 dB in the 200 to 1000 Hz range.

Have you set any type of low frequency level goal, or is your goal more just to use a tiny box?

 

[sreten]

Does this mean that a LT can both "remove" a possible
100Hz bump as well as boosting the lower end?

Hi,

Yes, that is what it does. Typically you cancel the pole zero pair around
100Hz and replace them with an effective pair around 50Hz, Q = 0.5-0.6.
If you don't have a bump to cancel, its not an LT, its simply bass boost.

Bass boost can be shelving or a peaking high pass function.
Shelving suits sealed moreso, high pass suits vented moreso.

gds, sreten,

 

[head_unit]

Hi,

Yes, that is what it does. Typically you cancel the pole zero pair around
100Hz and replace them with an effective pair around 50Hz, Q = 0.5-0.6.
If you don't have a bump to cancel, its not an LT, its simply bass boost.

Bass boost can be shelving or a peaking high pass function.
Shelving suits sealed moreso, high pass suits vented moreso.

gds, sreten,

Exactly...in mathematical theory. In actuality, it is a bit of nonsense, because the drivers' response functions
a) do not follow theoretical filter math
b) vary according to volume level, heating, the weather etc etc.

It may kinda work, but it can't REALLY work. You might as well EQ the in-room response.

Most of all, just stuff the box and see how it sounds. Even the theoretical difference in response between say 13L (10L stuffed) and 52 liters is not really night and day. Depending on program material, associated equipment, and most of all the listening room and ears/brain, it may be fine.