So lets say you have 2 or 3 subwoofers each in its own chamber, you could tune each to different frequencies so all bass notes would play at a similar volume. I think you'd need to power each one on an independant amp to set the subsonic and crossover frequencys seperately.
How far apart should they then be tuned to transition fairly smoothly, not necessarily perfectly, through the low frequencies?
I think that since ported enclosures are more effecient than sealed, you should in theory have similar sound quality and be louder than the same setup sealed.

I've been kicking this idea around for awhile now, are there any holes in my theory?

 

First of all, sealed boxes don't have a "tuned" frequency. A correctly designed sealed box will have a perfectly flat response above the -3dB frequency and roll off a 6dB/octave below that frequency. If done with ported boxes it's not really a good idea. Any time there's a rolloff in the magnitude of the frequency response there is also a phase shift. If you build ported boxes with different -3dB frequencies, the phase won't line up and you'll actually lose sound in certain frequencies. These phase shifts can be fixed with active all-pass filters, but that is VERY difficult and expensive to do. I'm not sure I could do it and I aced all my filters classes. Actually, it might even take digital filters to fix it, which makes it even more expensive. Essentially you'd need a computer-controlled filter setup to keep all your speakers in phase.

If you really want a flat response, use a sealed box of the correct size.

If you want to take a look at the relationship between magnitude and phase shift of a transfer function, look here:

http://users.ece.gatech.edu/~mleach...02/vcinduc2.pdf

This is an article on voice coil impedances, but it shows a phase and magnitude plot in a couple of places.


[Modified by GTLocke13, 9:07 AM 12/18/2002]

 

:crazy: I thought I knew quite a bit about car audio until I read that. :lol:

 

:iagree:

 

Sorry, I didn't WANT to point out that your theory has more holes than a cheese grater... :blueangel:

Acutally, you could make it work using 12"s or 15"s or even 18"s as subs and 8"s as midbasses. Cross over between them at around 80 Hz, then cross over your 8"s into your midranges at around 160 Hz. That setup wouldn't be too hard. But again, to do it right, you have to account for phase shifts in the crossovers.

If anyone is really interested in finding out how to do all this stuff, here's a link to the textbook I used for my audio engineering class. It was written by my professor and it is chock full of useful knowledge, from how to design simple crossovers to how to model entire speaker systems in pSpice. It pretty much requires understanding of complex equations (by complex, I mean dealing with real and imaginary numbers), vector calculus (a little), and transfer functions. But if you can get through the math, it is hands down the best audio design resource out there. If you don't have the mathematical knowledge to keep up, it can still be useful; the end of each derivation is usually a pretty simple, useful equation. But you might be better off with a simpler, cheaper audio design book.

http://www.amazon.com/exec/obidos/tg/detail/-/0787278610/qid=1040252945/sr=1-2/ref=sr_1_2/104-1116400-2338347?v=glance&s=books

Fixed, I think.


[Modified by GTLocke13, 6:17 PM 12/18/2002]

 

Wish your link worked in the first post!- I just get error messages. I agree- this is acoustic wave theory guys. If you have 2 speakers out of phase, they actually cancel eachother out at their similar frequencies. This is one reason (there are lots) why on infinite baffle speakers, you have to isolate the back of the speaker from the front. There IS a method of speaker design though, where you soothen out the peaks of a ported box using more than one tuning freuency, but they have to be a specific frequency/ octaves apart in order to produce a good response curve without what they call "nulls" in the response curve.-J

 

Fixed

 

Well those frequencies are what I want to know. Would you guys stop beating around the bush already! :D

 

I think he means using two ports in one box, not two different ported boxes. I haven't studied any systems like that so I couldn't begin to tell you how to go about it.

 

It's REALLY hard to do this without a proper computer program and REALLY precise tuning- but in one of the books I have it describes porting the enclosure with 2 ports tuned at 2 different frequencies one octave apart (I think...this is a reach here). What results is 3 different frequencies where there is a small peak in the response, but overall this is smoother than most sbb4 enclosures with a high Q- (You can tell these enclosures easily- the boom cars that have one boom frequency you can hear a block away- but higher or lower than that- it's almost inaudible). There is a rise in output at the first and second port-tune frequencies, and then one that is from the two ports together, which act as one large vent with the combined volume of the two ports. May be more trouble than it's worth though- epecially if you want to put it in a vette. This type of enclosure would NOT require much power at all- and would be quite large, even for an 8".

 

aka, not worth your time! Stick to sealed or ported. :yesnod:

 

Ports don't act as volumes of air. The important thing about a port is the ratio of length to cross section. You can increase the length as long as you increase the cross section. I forget what the acoustic term for the air in a port is, but the electro-acoustical equivalent of a port is an inductor. For a volume of air, the equivalent is a capacitor. This explains why ported enclosures have a resonant frequency and sealed enclosures don't. An RLC circuit has a resonant frequency where the impedance of the capacitor and inductor cancel each other out. An RC circuit doesn't have a resonant frequency. The book I posted above explains more about electro-acoustic modeling.

 

No, no, no, I mean seperate boxes tuned to different frequencies in the same vehicle.

Hey Jason I really wanna check your car out sometime if at all possible.


[Modified by korvetkeith, 5:48 PM 12/19/2002]