6th order series-tuned bandpass theory

[Triticum Agricolam 15]

I have a great interest in the science of box design and I've seen several examples around the 'net of folks building high-power series-tuned 6th order bandpass boxes for SPL purposes.  One common design theme seems to be to purposefully undersize the rear port, the reason given is for "cone control".  This appears counter-productive to me because the frequency where port velocity will be the highest, and thus having an undersized port will have the most effect (whatever that effect may be), is at the tuning frequency when cone excursion is at a minimum anyway.  It seems to me all these guys are doing is making a 6th order bandpass that becomes a 4th order basspass with a leaky rear chamber when they turn the volume up.  Basically, I think they are doing it wrong.  Does what I'm saying sound reasonable, or am I missing something here?

 

I would love to know exactly what's going on with these boxes.  If anyone has, or has a way of getting, a high power impedance sweep from one of these boxes in question that would tell me really quick what's happening.  A frequency response curve would be interesting to see as well.

 

I also strongly suspect that these series tuned boxes are not tuned to what the designers intend them to be, but that's another topic.  

 

Anyway, I look forward to anyone's input on this.  I probably should just build a test box and experiment with it myself, but until I find the time to do so I'd love to hear what folks have to say.

[SpeakerBoy 64]

Do you know how to model enclosures?

[Sencheezy 689]

Do you know how to model enclosures?

He's a very intelligent person. He knows how to model, hence his wordings. He's asking WHY do people in high SPL application apply this design, when by paper, it seems counterproductive. 

[SpeakerBoy 64]

What I meant by that was, firstly how is that different from peoples silly ratio assumption for 4th order bp's? And, if you are good with modeling, and have the spare wood, try out one versus the other.

I'm not sure what I'm doing when it comes to modeling, haven't spent time on a computer to figure it out yet.

[Triticum Agricolam 15]

Thank you for the compliment Sencheezy.

 

Yes, I do know how to model.  The issue is its a very difficult question to answer with modeling.  While I consider myself to be quite proficient with WinISD and Hornresp, neither of those applications can properly model series-tuned BP boxes.  Also neither can account for port compression.  I'm in the process of learning to use Akabak and have played around with some series-tuned 6th order BP designs in it, but I don't think it can account for port compression either and I'm not good enough with it yet to get any data out that I trust.

 

I think making the rear port undersized is probably a lot like people's silly assumptions regarding 4th order BP ratios.  I understand why those are silly, I want to understand if what people are doing with 6th order BP designs is silly as well. 

[///M5 2,833]

What makes you think there is possibly any merit?  As you are aware, the amount of confusion in car audio is horribly absurd.

 

If you go model it in Sysnoise you'll have your answer...but you obviously know it already.

[Triticum Agricolam 15]

What makes you think there is possibly any merit?  As you are aware, the amount of confusion in car audio is horribly absurd.

 

If you go model it in Sysnoise you'll have your answer...but you obviously know it already.

 

Well, like you said, I'm pretty doubtful there is any merit to what I've seen people do.  These same people seem to be very successful in other aspects of their designs and I figured I'd give them the benefit of the doubt until I have more absolute proof of what's going on.

 

Sysnoise looks like some very interesting software, though I have a feeling it is probably a bit out of my price range.  

[///M5 2,833]

Most companies that use Sysnoise invest 100's of thousands of dollars in it per year. It will pretty much model anything. That being said, I make a living out of verifying data from it and proving that it still needs experimental support (LMS/Siemens would agree btw). For what you want to model it would do it rather accurately and easily though. Complex multi-helmholz resonator design for intake manifolds and such it is great at.

As for the merit you question, I can't back it up with math (haven't tried either as it isn't worth the effort to me) but am absolutely sure you are correct. As you know random trial and error in designs couldn't be more idiotic. I can guarantee that is random as well.