Too much port area?

just a guy

Member
2006-05-12 6:59 pm
Bigger is better to keep velocity down but long ports can introduce problematic resonances so you have to use the proper software to find out if it's a problem or not.

Hornresp, Akabak and this one
http://www.diyaudio.com/forums/software-tools/220421-transmission-line-modelling-software.html
are all free. MJK's worksheets are good too, but there's a modest $25 fee to purchase those.

When the ports get long WinISD is not the right tool for the job, regardless of whether the new version claims to be able to handle port resonances or not.
 

just a guy

Member
2006-05-12 6:59 pm
On second thought, only Akabak and the linked software will simulate the weird ABC thing you are doing, so never mind the other two.

As an aside, a transmission line is essentially a huge port with the driver attached to the end (or side) so there's no such thing as too much port, only a point of diminishing returns and/or too much total volume.
 
Hornresp will do it.

Simulate as a vented box, take the port area to ~driver Sd, check the displacement graph to increase port length to get the same tuning as before.

An awful lot of resonances appear further up the pass-band. Not recommended if you want anything other than a traditional subwoofer (crossed over ~80Hz)

Chris
 

just a guy

Member
2006-05-12 6:59 pm
I'm pretty sure he's doing something like this...

5351600016_large.jpg


... and I'm pretty sure Hornresp won't do that.
 
AkAbak's learning curve is an obstacle for many. I've found the easiest way to run it in 64-bit Windows (assuming you have something other than Home Edition) is Microsoft's own "Windows XP Mode" VM. It's much less hassle to set up than a copy of XP in VirtualBox.

Looking at the picture above, that might be a DCR (Double Chamber Reflex). If so, it can be modeled using standard ported box software:

DCR
 
AkAbak's learning curve is an obstacle for many. I've found the easiest way to run it in 64-bit Windows (assuming you have something other than Home Edition) is Microsoft's own "Windows XP Mode" VM. It's much less hassle to set up than a copy of XP in VirtualBox.

Thanks for the tip.

Looking at the picture above, that might be a DCR (Double Chamber Reflex). If so, it can be modeled using standard ported box software:

DCR

That article states that a dcr can be made by rule of thumb and if done so it's ok to simulate it as a regular ported box. I don't think that's a good idea on both counts.
 
Provided that you stick to the design constraints (chamber volume ratios, port sizes) specified on Claudio's DCR pages, you can use a standard ported calculator. I've compared a standard ported enclosure and its derived DCR equivalent in AkAbak and it does work.

I'm sure it does work to some extent but even the author of the article says it has unpredictable aspects.

A typical frequency response of a DCR speaker has a dip above F3 that, according to Augspurger and Weems, is not to worry about since it's of modest amplitude and also because dips are less audible than peaks. In case the dip is very deep, Weems suggests to place some absorbing mat inside P2; however doing so will increase the port losses altering the known parameters in a not easily predictable way.

I prefer to simulate accurately and not have to worry about problems after the fact. And the suggested "fix" is stuffing the port, which kills the output.

Also, as we all know, as enclosures and ports get large they start to produce resonances inside the passband. I'm sure the articles' rules of thumb work fairly well for small boxes but when you get into large subwoofers the physical dimensions can start to cause big problems and there's no way that I know of to make sure it's not a problem other than accurate simulation. (You can build, measure, burn it, repeat, but I'm not a fan of that process.)
 
Last edited:
It's not my first build. The point is to get a really flat response, without overly sacrificing output. The low and high tuning cover a wider range than standard ported.

I'm doing this with two woofers, so I think a horn would be way to big for my trunk.

Thanks for all of your help though

The low and high tuning do not cover a wider range than standard ported. The high tuning simply provides a modest reduction in excursion and distortion up high (up near 100 hz in the article) where you don't really need it.

The bandwidth of a ported box goes as high as the driver can go (unless you have problematic resonances, which is actually MORE likely with this design - the article even says so in the quote in my last post). You can't get any more wide range flat response than a simple ported box.
 
Last edited:

OscarS

Member
2011-01-02 10:44 pm
I agree with the above post. The additional tuning frequency on the "high" end does practically nothing since excursion is already reduced in that vicinity, and then PLUS you have to deal with a dip that may or may not be audible, depending on the implementation.

Here is the proper Akabak script, using a Stroker15 (blue line is standard ported box of equal volume& tuning, overlayed)
abcII_zps88b8cf9f.jpg


Here is the excursion
ABCIII_zpsc241a2c4.jpg


As anyone can see, they are practically identical, save for the huge dip in the response (so much for trying to go for "flat" response, eh?), and the unsubstantial reduction in excursion in the upper bass frequency.

So, nothing has changed since I posted the other script in your "xmax...." thread. If you still don't know how to interpret frequency response graphs and excursion graphs, then you are handicapping yourself because proper simulations can lead to many eye-opening revelations about what works well, and what doesn't in enclosure designing.
 
Last edited:
Thanks for posting that, it gives a clear visual of how these things work - much more clear than the few words I posted.

If you still don't know how to interpret frequency response graphs and excursion graphs, then you are handicapping yourself because proper simulations can lead to many eye-opening revelations about what works well, and what doesn't in enclosure designing.

That's very sage advice and I'll take it one step further - with enough experience you start to understand how things work without even having to fire up the simulator.