Erin's 2006 Civic Sedan

[chad]

subs in the backseat floorboards, vent it forward and up behind the dash. one bigass bandpass.... well.. two.

[erinh]

there once was a time when I would have done that.

[claydo]

Multiple 5" drivers in a folded horn up firing through the dash?

Lmao, call it the Bateman special?

[chad]

there once was a time when I would have done that.

Do you use the back seat?

[erinh]

alright, some computer time...

Before I built these kicks I determined I could get roughly 0.30 cubes with the woofers pushed pretty far back and taking only the corner area like my JL setup did. But, knowing that would be too small I went with what I have now. At the time I considered doing an Linkwitz Transform (LT for short) but I wanted to give the whole floor enclosure thing a go. After 'giving it a go' and saying 'oh, no no no', I was back to the drawing board. Now, of course, I can always fall back to a setup similar to what I had before with the subs behind me but TBH the difference in tonality having all the drivers up front was a pretty noticeable improvement. One I really liked. Besides, where's the fun in that? I know I can make that sound awesome... I need a challenge. So, the LT idea is back on the table.

For those who don't know, an LT basically allows you to use a woofer in a small enclosure and create a filter (the LT) that reverses the high Q nature of the enclosure to the degree you desire. The drawback is you have to have a) power and b) excursion. Why? Well, because the filter is boosting the low end and you need unclipped power and excursion. How much power/excursion is a function of the transform. Note: This is NOT what a boosting the low end with your parametric EQ is: an LT is a filter that increases the low end, sure, but it does so in the same phase and magnitude that the larger (lower Q) enclosure would have had. You would need multiple EQ filters to achieve this... some pushing the low end up and also pulling the Fs area down (because, remember, a high-Q enclosure will exhibit a peak and then sharp rolloff; you're reversing this with an LT). With all those EQ filters you're going to do some crazy stuff to the phase and you won't achieve the exact transform needed. i'll give examples below. And you can find a LOT of information on LT's if you google. This definitely isn't new. It's just not seen in the car audio world. I honestly can't think of anyone who's actively using it (I have suspicions why, but who really cares).

How do you do a real LT? Well, miniDSP makes a plugin, so that's how I'll do it. I picked up a revB miniDSP with the DC isolator for $80 shipped. Score.
There are other means... one being building a circuit. But the miniDSP plugin makes it a lot simpler.




How do you determine the LT values for the miniDSP and how in the world does that relate to the enclosure/woofer you have? I'll explain...

First, let's take my Illusion C10 woofer and the 0.30 cube enclosure I have and calculate the basic "need to have" parameters: Qtc and Fsc (the Q and Fs of the sealed, closed enclosure). Pull up WinISD Alpha and get to work...
Modeling this yields a Qtc = 1.010 and a Fsc (Fs of the closed enclosure) = 73.58 Hz.

*I also used Bass Box Pro to model the effect of cabin gain (using 70hz as the +3dB Frequency with a 12dB/octave slope). This part isn't needed but I did it out of curiosity...

Small Enclosure (0.3 ft3) in Red with cabin gain in Orange:


Some might say this is just fine. But you're still going to have to deal with the high-Q ringing of the enclosure which is no bueno. Besides, I don't want a flat response in my car. So, on with the LT design.*


Second, download this excel sheet. This will tell you the LT biquad values (fancy way for saying filter) you use to punch in to the miniDSP. Additionally, this excel sheet will tell you how much 'gain' you need at your amp.
Use the Qtc/Fsc values you obtained from your baseline model. Punch those in to the excel sheet under F(0) and Q(0). Now, this is where you have to play and this is where I suggest pulling up winISD Alpha again. WinISD Alpha has the ability to model an LT. Doing this will give you an idea how much excursion you're running in to and just how well the LT design you want is going to target a desirable sealed enclosure emulation. This is what I did here...

Red = Woofer in 0.30 ft3
Blue = LT with a targeted fo of 40hz and a targeted Qtc of 0.50.
Yellow = Same as blue above but with 6dB BW filter @ 20hz to limit excursion




All compared with 300w:



300w with excursion (red horizontal line is marked at 10mm, but woofer xmech is 25mm):




With the 300w models, what you see above is the Red (baseline) never exceeds excursion but it has a very high Q. The Blue is the LT with a targeted Qtc = 0.50 and corner frequency of 40hz. The excursion is up there... I'd be leery of this. So, take the same LT and slap a 1st order butterworth filter on it at 20hz to limit excursion and you have the response in Yellow. This cuts down on excursion considerably below 30hz and if you look at the SPL curves, at 20hz the highpassed response is only 2dB. Additionally, from the excel sheet, it appears I'll need +9dB of gain. That's perfectly acceptable because I have enough power to cover this; currently my midbasses are cut about 10dB vs the midrange. So, all in all, the LT with a 1st order BW filter is a great compromise.

Now that I did all that research and modeling, I have a good idea of what I can expect the woofer to do in the given enclosure with a given LT in the miniDSP. It looks more than feasible to achieve what I want here. Of course, the proof will be in the pudding but doing due diligence up front is much better than throwing caution to the wind and finding a failure which modeling would have shown and saved me time on.


Also, remember, you can do this passively if you want. Here's a good link on how to build the network, once you've experimented and determined the ideal LT. That said, I still would recommend buying a miniDSP. The parts and time you'll spend piecing together a passive LT are probably not worth the trouble unless you're just in to that sort of thing. And IME, LT's do require a little bit of tweaking. Not major. But a little bit. It'll be much harder to do that without the miniDSP or some other electronic filter designer.



Something to remember here: since you're creating the response you desire by basically flipping the baseline response on it's head, you need to watch your power requirements and excursion. Coincidentally, remember that if you're targeting an LT Qtc = 0.70, it will require more power/excursion than a Qtc = 0.50 because a lower Qtc will roll off more gradually where a higher Qtc will have a sharp 'knee' rolloff; meaning that your LT would be trying to make this sharp knee but also extend the rolloff frequency further down before said knee.

[erinh]

Let me note the above models are really only rough cuts at what I think I want to do. So, don't take it as literally as it may appear. The targeted F(p) and Qt(p) I use may very well change a bit. But like I said above, this is where having the miniDSP pays dividends: you can make these adjustments on the fly to get a real idea of the performance vs what your needs are.



And, just to reiterate, the LT idea would allow me to go to a kick panel install where the woofers are at the corners of the kicks and the enclosures are basically just large enough to house the woofers. No more floor enclosure on the driver side and the passenger side enclosure would be cut off at the firewall.

As it is, the passenger's side enclosure imparts practically no vibration through the car thanks to 'decoupling' it via the Extruded Butyl Rope from SDS. Being able to A/B between the two kick panel's side, I determined that using EBR again in the same fashion on both sides should be adequate to keep from having tactile at my feet on the driver's side. It's simply a function of getting my feet of the dang enclosure.
Well, in theory, at least. Guess we will see.

[BigAl205]

tl;dr










It think that's a cool idea to try out. My only concern is that you will still get transfer from the enclosure to the floorboard, and eventually to your foot. Since I never got to hear it, is the vibration you feel directly from vibration, or from the pressure wave?

[claydo]

Ok, I think I understand.......sounds like yer gonna need a case or two of spare c10s, lol. Is your current power gonna be good to cover this plan? Does your modeling give ideas of the power needed to meet the demands of the LT filter? This sounds risky, yet the benefits of working a woofer into available space, and shaping the response, yup, better have some spares.........

Hmmm, will pushing the c10s excursion affect it's ability to reach up cleanly to the 2.5s? I guess the better question would be, when mixing high excursion with wide response, is distortion the only spec that could alter the tonality up in the higher response area of the driver?

[erinh]

Clay, I think all your questions are covered in my post. Short of the one about reaching up high. The answer to that is its no more a concern than it is with the current enclosures, whose QTC is roughly 0.50 or so. From modeling and comparing, it's not something I'm worried about if I go with the plan I mentioned above.

I think you may be overstating the need for spares. I don't have that kind of money to replace these woofers so if I really were concerned with thermal handling then I definitely wouldn't be going this route. Lol.

[claydo]

Lol, I know I was exaggerating....for dramatic effect! I just thought this sounds dangerous to woofers.....as long as the modeling algorithms are correct for this process, I'm sure you'll he fine!