Small, Sealed Box for Sub; Linkwitz Transform Circuit

[wrestlinguy]

I know what you are thinking..."A sealed box for a Sub, you've got to be kidding me. Is this guy for real?" But with the help of the Linkwitz transform circuit found on Rod Elliot's webpage, http://www.sound.westhost.com/project71.htm , it is possible. However, I do have a quesiton about using this circuit.

Rod does an excellent job of outlining how the circuit works and how to select the various components for your specific needs, but it appears to me that he is only considering the case where there is only one driver in the box. However, I was wondering if I were to use two identical drivers in the same box, would this effect the equalization of the system so far as to effect component selection? The attached spreadsheet on Elliot's site calculates the component values for you based upon speaker and box characteristics. How would the introduction of another driver into the box effect the inherent resonance of the box?

So I guess to make my question much simpler, aside from volume displacement, what other changes do I need to be aware of by introducing the second driver?

Thanks,

Shawn

[cocolino]

I know what you are thinking..."A sealed box for a Sub, you've got to be kidding me. Is this guy for real?"

Not at all - I do this all the time and it works great!
I do prefer this over any BR-enclosure.
Moreover this gives much more choices of types of alignments by simply changing the coefficients without building a new enclosure.


The calculations for the LR-transform circuit for a driver in a sealed enclosure are based on Otc (Qts) and fc (fs) where:
Qtc = Q_box (enclosure with driver)
Qts = Q_driver (driver only)
fc = f_box (enclosure with driver)
fs = f_driver (driver only)

With two drivers (presupposed they`re identical!) all You have to do is to double the enclosure volume.
Then the calculation and the component values for the LR-transform are the same as for a single driver with half the volume.


When You add a second driver (again presupposed it is identical!) in an existing box (with one driver), Qtc as well as fc increases (because Vas doubles due to two drivers).
I´m a little lazy now to present the formulars for calculating sealed enclosures but I`m sure You know where to find.
Do the equotations for a sealed enclosure again but now with double the Vas value (all other values as Vbox, fs, and Qts remain the same). As mentioned You`ll get a higher Qtc as well as a higher fc in this case.
Now, take these values for doing the LR-transform circuit equotations again.

BTW: better than calculating Qtc and fc is actually measuring them

[Pjotr]

Hi Shawn,

Fully agree with cocolino. No kidding at all In the past I experimented a lot with electronic bass correction for CB and I have had great results with it. HERE is a 415 kb .zip file with a short engineering note about it. There is also an Exel spreadsheet to calculate a filter and to plot the box and filter responses with a given speaker unit in a desired target box. Although I used a different filter topology than the Linkwitz transform, the results are the same.

Cheers

[sam9]

Actually one the best deals arounf is the Dayton Titanic Sub kit @ $349. A sealed box sub. You can even down load info on how to rewire the built in EQ to modify the response curve.

So- I don't think anyone is likely to think you are crazy.

[Matttcattt]

i am thinking of building rod elliots project 71, and i have also seen project 48, the Sub-Woofer Controller, and i am not sure which to use. they both extend the low frequency. i am thinking of building both when i build my sub woofer. any advice?

[phase_accurate]

Although I used a different filter topology than the Linkwitz transform, the results are the same.

This is not exactly true but with a tendency to the better side: Yours is less restricted in terms of possible alignments.

The only disadvantage is component-count.
Another possibility would be the use of the "classic universal filter" topology but this would be a little less convenient to tune than yours.

Regards

Charles

[Pjotr]

Hi Charles,

I’ve fiddled a lot with the LW transform those day with doubtful results. It lacked the possibility to align and it tended to be rather sensitive to component tolerances.

So we took the way as outlined in the note. Although all alignments are possible, even tuning the new Fs_box far above Fs of the speaker unit, the maximum filter boost is restricted to about 10 dB for real speakers and amps.

Although Fliege filters take 2 opamps and more components they are more easy to tune than standard filters and tuning components are the same for the 3 filters required. Also they are more tailored to low-Q filters. The extra component count outweighs the hassle of finding the right component values for a LW transform. Also it is not so difficult to find 3 closely equal capacitors. I have used the filter also for mid speakers to tune them as half part of a 4th order LW x-over between bass and mid.

Shawn,

The spreadsheet attached to the note does provide for more (equal) units in one enclosure, giving you the total Fs_box and Qt_box, also when you want to use the Linkwitz filter. It does not take into account eventually used stuffing, measuring the actual box for final tuning stays the best approach. If you want to use stuffing material you can fill in a 1 to 1.2 higher box volume then what you actually are going to use.

Cheers

[SY]

So I guess to make my question much simpler, aside from volume displacement, what other changes do I need to be aware of by introducing the second driver?

Two things to keep in mind:

1. You HAVE to measure the driver combination you have in the box you have. Using datasheets will only get you randomly good results.

2. You HAVE to be aware of the limitations of your drivers, excursion and thermal.

With the right drivers, a realistic notion of what you're going to accomplish, and taking the time to do it right, the LT can work very, very well.

[tiroth]

What is a reasonable goal to shoot for with LT? Looking at some of the adire plots, it looks like a corner-loaded sub can easily be flat to 20Hz with only 6dB of gain. Throwing that out the window, what should we really shoot for? Room gain IS going to be a factor, regardless of who does the math. So if we are concerned only with the 20-40Hz octave, how flat should the response really be?

I'm guessing that with LT the usual rules apply; that is, Qtc < .707 is desirable. This means Fsc must be very low indeed if Qtc is around 0.6 and we want flat to 20Hz...but do we really? How far down is acceptable given boundary effects?

I guess that my goal would be to avoid putting a lot of effort (and power) into getting the response flat to 20 only to discover that loading actually produces a hump in the response. FWIW, I would probably be locating such a sub between my towers, so it would only be loaded by one wall.

I can already see that despite the power of LT, a lot of iterative testing is needed. At least it doesn't require building a new box!

[cocolino]

originally posted by Pjotr
I’ve fiddled a lot with the LW transform those day with doubtful results. It lacked the possibility to align and it tended to be rather sensitive to component tolerances.

Hello Pjotr,
what exactly do You mean with "lacked the possibility to align"?

I parallel R`s and C`s to get the sometimes odd values required and with 1% metal film resistors (quite normal) and equally precise matched C`s, measuring a couple of my LR-Transform circuits showed that in practise they work almost ideally according to what they should do theoretically.

I mean every audio electronics DIYer should really have a LCR-meter for matching parts. Even the affordable handheld meters are accurate enough for this.

originally posted by SY
2. You HAVE to be aware of the limitations of your drivers, excursion and thermal.

This is VERY true. With the possibilities of the LR-Transform one might be tempted to force drivers to what they are not suitable.
But not only the drivers have their limitations, also the power amplifier has to be much stronger than without LR-transform.
Think of it: for example extending the low frequency range by one octave below fc requires 12dB voltage boost which in turn equals 16-times power!
Finally the LR-circuit itself is susceptible to overdrive and voltage clipping when the low frequency extension is high or even overdone.
Not only because of this IMO a highpass (subsonic) filter preceeding the LR-Transform is a must. Some program material can contain weird ultralow frequencies. This hasn`t necessarily to do with instruments as organs for instance. I remember having a CD with only two instruments - a female voice and an acoustic guitar. This is hardly what one thinks can produce dangerous low frequency levels but the singer is very close to the microphone thereby producing some kind of ultralow frequency winds - I have never seen my two 30cm long throw subwoofer drivers jump so much (and this is WITH a subsonic filter already).
Well admitted, my subs are tuned to very low frequency and it`s an extreme case but You should be aware of those effects when using the LR-Transform circuit (or any other kind of low frequency boost circuit).