Looking at the variety of commercially marketed active subwoofers from all the manufacturers, it is quite apparent that most of them shrink the box down to the smallest practical size. A typical unit might use a 12 inch driver in a 40 liter sealed box. Through the use of equalization, the resulting sub has (or is claimed to have) a response down to 20 Hz.
Are there any practical guidelines on how much equalization can be tolerated before sound quality degrades? Are there other considerations or trade-offs to shrinking a subwoofer box down to well-below optimum size?
As an example, the SB Acoustics SB34NRX75-6 12" woofer would seem to work well in a 100 liter sealed box (Fs=19 Hz, Qts=0.4, Vas=260, Xmax 0-peak= 11mm). If this box were shrunk down to 50 liters, the resulting system would need a boost of about 6 dB at 25 Hz and a cut of 1 dB at 70 Hz to match the optimal 100 liter box. What are the down sides to shrinking the box and doing the equalization (other than the need for DSP and plenty of power)?
Thank you for your thoughts
Are there any practical guidelines on how much equalization can be tolerated before sound quality degrades? Are there other considerations or trade-offs to shrinking a subwoofer box down to well-below optimum size?
As an example, the SB Acoustics SB34NRX75-6 12" woofer would seem to work well in a 100 liter sealed box (Fs=19 Hz, Qts=0.4, Vas=260, Xmax 0-peak= 11mm). If this box were shrunk down to 50 liters, the resulting system would need a boost of about 6 dB at 25 Hz and a cut of 1 dB at 70 Hz to match the optimal 100 liter box. What are the down sides to shrinking the box and doing the equalization (other than the need for DSP and plenty of power)?
Thank you for your thoughts
Last edited:
None really.* It is not common in the DIY world, because traditional speaker simulation software spits out 'maximum flat' alignments and people tend to like flat graphs and forget the room the speaker is placed in.
* Actually, the woofer must have a high power voice coil to cope with the extra amplifier power, have a strong motor to be left with some efficiency etc. which makes it more expensive. Distortion slightly increases because air is a non linear spring, which primarily adds 2nd harmonics. Furthermore, distortion increases because voice coil inductance varies with current for low to medium cost (or old design) woofers, while woofers need more current when mounted in a small box. This also heats the coil, which increases its series resistance and modifies the frequency response. Motional feedback fixes these distortion and frequency response issues, at the cost of complexity.
Last edited:
That is the impression I got from Floyd Toole's book. If a woofer is put in a non-optimal box and then equalized to be flat, it is sonically equivalent to being flat from the stand point of frequency response and group delay. That is the theory.
Has anyone done it both ways and compared them?
Has anyone done it both ways and compared them?
Due to the increased power level, there will be a lot more flux modulation below Fcb, so the distortion performance will be a lot worse. The time constants created by shorted turns will not be long enough to deal with this, so even a low distortion woofer in the 100Hz range, will have much worse distortion below Fcb. It is important to remember that since the box is really small Fcb is much higher than normal, so the bandwidth over which the distortion level is higher than normal is extra wide.
Intuitively I suspected that the increase in current through the voice coil (for a given SPL) must somehow lead to more distortion, but I don't know enough about electromagnetics to describe why. Thanks.
Keeping with the above example, the displacement limited acoustic power of the SB 12 inch driver is shown in the attachment. At 20 Hz, the driver runs out of x-max at 99.4 dB, which requires just 13.7 W. So a mild boost of +6dB at 20 Hz would raise the power level to ~ 55 W, which seems well within the capability of the driver. Of course the acoustical output of the driver is limited to 99.4 dB SPL regardless. But anything over +9 dB would seem problematic.
Thanks for your thoughts.
Attachments
That fig 11.6 gives one a very inaccurate view of what goes on with a driver in two different boxes.
There are two springs in parallel in this situation. One is the driver's (spider+surround). The other is the air in the cabinet. The first spring is quite nonlinear. The second spring is more linear than the first, even with a small box. The result is that the second spring dominates the behavior more and the system behaves more linearly. You can model this as nonlinear parallel resistors (R1+R2/R1*R2).
There are two springs in parallel in this situation. One is the driver's (spider+surround). The other is the air in the cabinet. The first spring is quite nonlinear. The second spring is more linear than the first, even with a small box. The result is that the second spring dominates the behavior more and the system behaves more linearly. You can model this as nonlinear parallel resistors (R1+R2/R1*R2).
IMHO the validity of your assertion is dependant of the loudspeaker driver contruction (the membrane should be stiff and heavy and therefore the spider as well). But it would be a good idea for a loudspeaker driver that is specifically designed to be tightly enclosed in a tiny box to have a spider geomety that induce a little less stiffness in one direction ? no ?
😀
Last edited:
An interesting and learned discussion.
Funny, I thought it was an accepted article of faith that the air in sealed boxes, if isothermal (or if adibiatic?) leads to more linear results, at least since 1955?
In any case, distortion in the lower bass is not much of an audible problem and may not be important to fret about it for the amounts increased - at least theoretically - by a small box.
Likewise, for most people who have posted on the matter, flat doesn't sound as right as boosted bass - so the benchmark is bass boosted and that may take lots of watts.
Motional feedback is, once again, the answer. It has often been associated with getting bigger bass from smaller boxes along with fixing the FR by the magic of negative feedback.
B.
Funny, I thought it was an accepted article of faith that the air in sealed boxes, if isothermal (or if adibiatic?) leads to more linear results, at least since 1955?
In any case, distortion in the lower bass is not much of an audible problem and may not be important to fret about it for the amounts increased - at least theoretically - by a small box.
Likewise, for most people who have posted on the matter, flat doesn't sound as right as boosted bass - so the benchmark is bass boosted and that may take lots of watts.
Motional feedback is, once again, the answer. It has often been associated with getting bigger bass from smaller boxes along with fixing the FR by the magic of negative feedback.
B.
The biggest problem IMO is that such under-sized boxes tend to be closed, and it does not matter much how fancy your woofer is. More power to compensate for lack of low frequency SPL. the hot air cannot escape effectively.
Probably better to use alu cones in such boxes.
Probably better to use alu cones in such boxes.
A very valid option, unless there's others in the household that have preconceived ideas about aestethics.
I'm not sure if aesthetics is the most important thing to worry about when clobbering a subwoofer with a KW of power in a home setup 😀
You one are of the few people that is still thinking that a fine submilimeter absolute precision can be accuratly captured with a velocty information only.
Really 😀
That didn't stop the Bag End Infrasub line from being very popular and well regarded, and it operated completely below Fb, which was around 100Hz I think!
Or perhaps that is just a testament to just how much distortion most people can find "acceptable".
I have come to the conclusion that for any high performance closed box sub you want the box to be about as large as is practically possible. This maximizes the efficiency at the lowest frequencies and therefore minimizes power (current) demand for a given SPL. What's not to like, except the fact that you can easily dream up something that is large and therefore so heavy you cannot move it!
If you are concerned that the suspension non-linearities will be too large, then do some feedback/correction, e.g. employ a balanced bridge or whatever.
If you are concerned that the suspension non-linearities will be too large, then do some feedback/correction, e.g. employ a balanced bridge or whatever.
Last edited: