I've seen subwoofer designs where the port has a very restricted center section, and gradually tapers towards both ends. Sometimes they have an almost elliptical flare at the exit. I don't know what they are called, so if you know what they are called, that'd be a big help!
For the record, I don't mean a straight port with flared ends.
Examples are Genelec and their LSE subs, and I have seen pro audio speakers that use ports like this also.
On the Genelec, I suspect the restriction in the middle of the port reduces turbulence, since the port follows an arc. And on the pro audio side, I think maybe it allows a shorter port length while minimizing port noise/ resonance?
I'm asking because I built a radius port awhile back, and now I am interested in trying to improve the design. But I was just guessing, and I don't know the theory behind such a port.
The port I designed has a 10% restriction in the center. To determine port length, I made the inner and outer surfaces the length that VituixCAD calculated. (I used the length of the surfaces based on the realization that air flow will "stick" to those surfaces.)
My assumption about a radius design is that there will be turbulence as air reverses direction rapidly while trying to follow an arc. So my intuition says that restricting the port cross section in the middle section compresses the air, reducing this turbulence.
Questions I'd like to answer:
-Are my thoughts even remotely close to being accurate?
-How much restriction can I add before hurting the port's output?
-Do I need a specialized tool to calculate this type of port, or can I start with a calculation for a straight port?
-To achieve a desired tuning frequency, do I have to keep the port's internal volume the same as what a calculator says?
-Or does the restriction in the middle of the port offset the reduction in internal volume?
I will probably do some experiments at some point, but I don't know when that will happen. I'm in the middle of a long term project that deserves my full attention, so if you see me doing subwoofer testing, you will know that I am being irresponsible... 😛
For the record, I don't mean a straight port with flared ends.
Examples are Genelec and their LSE subs, and I have seen pro audio speakers that use ports like this also.
On the Genelec, I suspect the restriction in the middle of the port reduces turbulence, since the port follows an arc. And on the pro audio side, I think maybe it allows a shorter port length while minimizing port noise/ resonance?
I'm asking because I built a radius port awhile back, and now I am interested in trying to improve the design. But I was just guessing, and I don't know the theory behind such a port.
The port I designed has a 10% restriction in the center. To determine port length, I made the inner and outer surfaces the length that VituixCAD calculated. (I used the length of the surfaces based on the realization that air flow will "stick" to those surfaces.)
My assumption about a radius design is that there will be turbulence as air reverses direction rapidly while trying to follow an arc. So my intuition says that restricting the port cross section in the middle section compresses the air, reducing this turbulence.
Questions I'd like to answer:
-Are my thoughts even remotely close to being accurate?
-How much restriction can I add before hurting the port's output?
-Do I need a specialized tool to calculate this type of port, or can I start with a calculation for a straight port?
-To achieve a desired tuning frequency, do I have to keep the port's internal volume the same as what a calculator says?
-Or does the restriction in the middle of the port offset the reduction in internal volume?
I will probably do some experiments at some point, but I don't know when that will happen. I'm in the middle of a long term project that deserves my full attention, so if you see me doing subwoofer testing, you will know that I am being irresponsible... 😛
Not sure whether it is what you mean but I started a thread about port geometries, including "flared" ports with circular curved walls with narrow section in the middle of the port length.
I made an excel calculation tool based on research by salvatti/devantier/button:
https://www.diyaudio.com/community/...rbers-and-port-geometries.388264/post-7770554
The thread is quite long, but you'll find an index in the first posting. Or just ask me!
I made an excel calculation tool based on research by salvatti/devantier/button:
https://www.diyaudio.com/community/...rbers-and-port-geometries.388264/post-7770554
The thread is quite long, but you'll find an index in the first posting. Or just ask me!
I'll check it out. The Genelec port I was thinking about is this:
https://images.ctfassets.net/4zjnzn...e946d0422c70d66f320/lse_see-through.jpg?w=930
Now I'm noticing that it has a constant cross section for the majority of the port. I thought it had the gradual taper on both ends...
I'm not finding a picture, but I remember seeing a port design that was essentially two flares back to back with no straight section. But I'm having trouble finding where I saw that.
https://images.ctfassets.net/4zjnzn...e946d0422c70d66f320/lse_see-through.jpg?w=930
Now I'm noticing that it has a constant cross section for the majority of the port. I thought it had the gradual taper on both ends...
I'm not finding a picture, but I remember seeing a port design that was essentially two flares back to back with no straight section. But I'm having trouble finding where I saw that.
L'Acoustic's hourglass shaped laminar port might be what you were thinking of.
https://www.diyaudio.com/community/...h-push-slot-loaded-build.409197/#post-7602789
https://www.diyaudio.com/community/...h-push-slot-loaded-build.409197/#post-7602789
Welter, yes, that's the one! The link you used in that other thread appears to be dead, but I found a write-up by the company that wrote the modeling software. Happily you copied a few more diagrams into that thread than what is in the article below:
https://www.comsol.com/story/the-si...sure-design-on-loudspeaker-performance-102111
For the last few weeks I have been imagining a test cabinet that I can attach ports to the exterior of to test more easily. Looks like I'm not the first one to think of it.
But I guess the fact that L'Acoustics needed custom software means that I'm kind of on my own?
https://www.comsol.com/story/the-si...sure-design-on-loudspeaker-performance-102111
For the last few weeks I have been imagining a test cabinet that I can attach ports to the exterior of to test more easily. Looks like I'm not the first one to think of it.
But I guess the fact that L'Acoustics needed custom software means that I'm kind of on my own?
Alright, this might be really obvious to everyone else, but I never made this connection before. The article I linked above says, "For example, a bass reflex enclosure design incorporates a vented opening called a Helmholtz resonator."
I never made the connection between subwoofer port and helmholtz resonator. I guess it is different because it is excited internally, but a helmholtz resonator calculator gives me the same results as cabinet design software.
This answers one of my questions. The critical element is the port opening cross section and the volume of the port. So changing the shape or the length shouldn't change the tuning of the port, as long as the interior volume stays the same.
But a new question--is the tuning frequency dictated by the internal port opening or the external opening? Or a combination of the two? I suspect it is primarily determined by the internal port opening.
Looking at the Genelec design--the internal port opening is rectangular with a small flare, but the external opening is quite large with a gradual flare...and there is a large cavity under the subwoofer expands the port exit even further. So if we calculated port tuning using the external port opening, it would give us a much higher tuning frequency than what the Genelec subs have.
Please comment if you have more insight or experience with this?
I never made the connection between subwoofer port and helmholtz resonator. I guess it is different because it is excited internally, but a helmholtz resonator calculator gives me the same results as cabinet design software.
This answers one of my questions. The critical element is the port opening cross section and the volume of the port. So changing the shape or the length shouldn't change the tuning of the port, as long as the interior volume stays the same.
But a new question--is the tuning frequency dictated by the internal port opening or the external opening? Or a combination of the two? I suspect it is primarily determined by the internal port opening.
Looking at the Genelec design--the internal port opening is rectangular with a small flare, but the external opening is quite large with a gradual flare...and there is a large cavity under the subwoofer expands the port exit even further. So if we calculated port tuning using the external port opening, it would give us a much higher tuning frequency than what the Genelec subs have.
Please comment if you have more insight or experience with this?
Last edited:
The helmholtz resonance is defined by enclosure volume (acting as oscillating spring) and the weigth/surface of the port air (acting as accelerated weigth).
Therefore for resonance calculation of a classic tube port enclosure the simulators need
- enclosure volume
- port length
- port surface
Furthermore the effective port length should include end correction factors, depending on port ends having flange/no flange/vicinity to walls. Also, air density and speed of sound will affect tuning, but for usual situations the standard values are perfectly ok.
For non-tubular, flared or otherwise shaped ports the transition from "air spring volume" to "air weigth" is more fluent and calculation is getting less straigthforward.
Very roughly you could say that the minimum diameter/cross section surface of a port is relevant for tuning. But as stated above it's not that straigth-forward.
It is a velocity ramp.
And if velocity is too high it makes noise.
18 to 25 m/s is accepted velocity
Larger opening reduces velocity.
Tradeoff is extended length
If too long= no fit in box.
Curving ,bending, folding etc etc
is one solution for fitting it in there.
Tradeoff is turbulence noise and efficiency
losses. Another solution is just make it too small
and accept the higher velocity
So there is numerous applications
AKA flares or dual flares to reduce exit noise.
When accepting higher velocity as a tradeoff.
AKA trying to weasel a smaller shorter tube into
a small box and reducing the unavoidable turbulence
noise from high velocity. Another issue is being to long.
Good example is excessive barfing of T lines.
Early detection of aircraft use to be exhaust noise.
So most understanding of tube velocity and error factor calculations
Came from Naval research a very very long time ago.
When exhaust noise was basically a death sentence.
For urban applications and cars, vehicle speed and
horsepower less a issue with restrictive systems.
That method is well known, a muffler.
AKA stuff the heck out of a tube with absorption
material. Ironically the same method for fixing
T lines.
Problem is port output/ efficiency is reduced and rather pointless.
Ironically the ideal shape for circular air flow is a circle.
For less restriction. Mind blowing.
Why correction factors for rectangles and triangles have been
figured out, being they are restrictive and cause additional noise/ friction
Friction in all basic physics is a = loss of energy
more specific is a transfer of energy and considered a loss
specific to certain applications.
Applied to absorption the loss is wanted, sound pressure is turned
into heat. Applied to a port the additional pressure is wanted
Making the tube smaller in the middle would reduce needed length.
And tradeoff is always the same = Higher velocity
And if velocity is too high it makes noise.
18 to 25 m/s is accepted velocity
Larger opening reduces velocity.
Tradeoff is extended length
If too long= no fit in box.
Curving ,bending, folding etc etc
is one solution for fitting it in there.
Tradeoff is turbulence noise and efficiency
losses. Another solution is just make it too small
and accept the higher velocity
So there is numerous applications
AKA flares or dual flares to reduce exit noise.
When accepting higher velocity as a tradeoff.
AKA trying to weasel a smaller shorter tube into
a small box and reducing the unavoidable turbulence
noise from high velocity. Another issue is being to long.
Good example is excessive barfing of T lines.
Early detection of aircraft use to be exhaust noise.
So most understanding of tube velocity and error factor calculations
Came from Naval research a very very long time ago.
When exhaust noise was basically a death sentence.
For urban applications and cars, vehicle speed and
horsepower less a issue with restrictive systems.
That method is well known, a muffler.
AKA stuff the heck out of a tube with absorption
material. Ironically the same method for fixing
T lines.
Problem is port output/ efficiency is reduced and rather pointless.
Ironically the ideal shape for circular air flow is a circle.
For less restriction. Mind blowing.
Why correction factors for rectangles and triangles have been
figured out, being they are restrictive and cause additional noise/ friction
Friction in all basic physics is a = loss of energy
more specific is a transfer of energy and considered a loss
specific to certain applications.
Applied to absorption the loss is wanted, sound pressure is turned
into heat. Applied to a port the additional pressure is wanted
Making the tube smaller in the middle would reduce needed length.
And tradeoff is always the same = Higher velocity
Last edited:
Yes, the velocity will be increased at the restriction. But is it bad for a port to have higher velocity in the middle, if the velocity is reduced at either end by gradually increasing the diameter/ cross section of the port?
I suspect that narrowing the port at a change of direction will increase velocity, but it will also increase efficiency by promoting more laminar flow. And I think it is the exit velocity that matters, not the velocity inside the port. It is just a hypothesis....
no that's not bad, you are absolutely right!
as long as the velocity stays below the point where the boundary layer will induce turbulence.
depending on the length of the port (longer is "better") the velocity should stay below an absolute maximum of 30 m/s.
however, this limit is only valid for a port with very good geometry, otherwise there will be chuffing and turbulence below that velocity, e.g. in case of a sharp edge at the port exit/entrance.
I collected data from several ported speakers and tried to find the relevant correlation between geometry and compression (which again correlates with noise and distortion):
https://www.diyaudio.com/community/...rbers-and-port-geometries.388264/post-7755603
the investigation is still ongoing!
As far as I found out and as described in different papers the relevant parameter is the air displacement in relation to the port (exit) radius, called "strouhal number".
But you are still quite right, because the air displacement is related to the air velocity, but also to the output frequency.
for details have a look at David McBean's explanation:
https://www.diyaudio.com/community/...rbers-and-port-geometries.388264/post-7697492
and my procedure to find a suitable port exit diameter:
https://www.diyaudio.com/community/...rbers-and-port-geometries.388264/post-7697685
while air displacement should be small for the port exit/entrance, it is quite ok to have a much bigger displacement and a smaller radius at the middle port section.
Last edited: