# of elbows in a port

[Paul Carmody]

Working on a sub. The port I'm considering is a 3" piece of PVC--however, it has to be 24"+ long! I have no problem doing one elbow, but if I did two of them, it would give me a lot more flexibility in regards to the overall shape of the enclosure. Is two elbows a problem? Anyone care to comment from theory or practice?

[Jeff B.]

Re: # of elbows in a port

Working on a sub. The port I'm considering is a 3" piece of PVC--however, it has to be 24"+ long! I have no problem doing one elbow, but if I did two of them, it would give me a lot more flexibility in regards to the overall shape of the enclosure. Is two elbows a problem? Anyone care to comment from theory or practice?

The problem isn't so much the elbows and their bends - they may add a small amount of port resistance and lower Qp slightly, but not a lot - the problem is the length. A length this long will produce some line resonance peaks at higher frequencies that may be an issue. I would make sure that the port exists at the rear or at the bottom of the enclosure to ensure that the these are not audible in the midrange.

[nikbrewer]

Re: # of elbows in a port

what frequencies would this be an issue at? i have doe longer ports on subs and never knew about this

[Paul Carmody]

Re: # of elbows in a port

The problem isn't so much the elbows and their bends - they may add a small amount of port resistance and lower Qp slightly, but not a lot - the problem is the length. A length this long will produce some line resonance peaks at higher frequencies that may be an issue. I would make sure that the port exists at the rear or at the bottom of the enclosure to ensure that the these are not audible in the midrange.

Yeah, I have noticed those resonances in Unibox when modeling stuff. I did this enclosure in your WBCD spreadsheet (tip o' the hat), and with a Low-Pass filter at 70-100 Hz, I don't see any harm with the resonances. What do you think?
Nonetheless, the port will be bottom-firing, as will be the driver. (This is for a friend and his wife, who just had their first baby last month, so I'm planning ahead for "exploring fingers")

[john trials]

Re: # of elbows in a port

The HSU Research VTF-3 with turbo charger has 3 elbows in each port (one elbow is in the main enclosure).

http://www.audiojunkies.com/product/...harger-reviews

[Jeff B.]

Re: # of elbows in a port

Yeah, I have noticed those resonances in Unibox when modeling stuff. I did this enclosure in your WBCD spreadsheet (tip o' the hat), and with a Low-Pass filter at 70-100 Hz, I don't see any harm with the resonances. What do you think?
Nonetheless, the port will be bottom-firing, as will be the driver. (This is for a friend and his wife, who just had their first baby last month, so I'm planning ahead for "exploring fingers")

I missed the word "sub" the first time. If you are crossing that low then you shouldn't have enough energy at higher frequencies to excite the line resonances. It should be OK as long as your port air speed is within reason.

Jeff

[---k---]

Re: # of elbows in a port

Jeff,

I've heard people claim that slot ports don't exhibit the same resonance properties of the round port. Is this true?

[Paul Carmody]

Re: # of elbows in a port

All right, since we're going down that path anyway...

Let's say I use a slot port (of equal or larger surface area). How many "turns" are detrimental in that case? Is a 180 degree turn "worse" than a 90" turn?

Sub experts, speak up!

[nikbrewer]

Re: # of elbows in a port

well i have seen t-lines with like 20 180 degree f9olds, so i would guess what ever fits. I know i have done several with quite a number of folds

[Jeff B.]

Re: # of elbows in a port

well i have seen t-lines with like 20 180 degree f9olds, so i would guess what ever fits. I know i have done several with quite a number of folds

What we need to remember here is that transmission lines are forms of vented speakers with slightly different characteristics. However, all transmission lines in addition to to having a quarter-wave (and other variable wavelength resonances) also have an Fb due to cabinet tuning. Conversely, all vented speakers with tube or slot ports will have some quarter-wave resonance.

What distinguishes between these two is that in the typical vented speaker the volume is large compared to the port, which is normally relatively short in length. As a result the resonance is dominated by the cabinet Fb and the QW resonance is quite high in frequency.

In the TL the cabinet volume loading the driver is typically smaller and the port is large and very long. As a result the tuned Fb is now relatively high in frequency and damped out as much as possible and the line length's QW resonance now dominates the output.

It is also possible to combine the characteristics of the two by maintaining a fairly large volume behind the woofer to load it, and also include a large and long port. As you do this, you begin to combine the resonances of the two. As these two resonances get close together, then you may have problems. This is where using fairly long ports in vented systems can become an issue. - Just something to watch out for. As Paul points out Unibox will model this for you.

Jeff B.

[Max_Andrews]

Re: # of elbows in a port

The shallower the turn, the better, as it will add less resistance. I know in the hvac world there is a rule that every 90 degree bend you add robs you of 25% of the airflow efficiency or some other high number close to that value. So two bends in a 3" port is only going to give you the airflow of a 2" port, making it kinda pointless.

FWIW, I've also heard that the closer to "square" a slot port is, the better. (e.g. a 3"x3"x10" slot port is better than a 9"x1"x10" slot port)

I'd say the solution is simple: Increase your box size!

[gregrueff]

Re: # of elbows in a port

I used to debate this often on this board.

There is an old thread where I tried to argue that a 90 degree bend would induce minor losses in the port and skew the tuning frequency. Bill Fitzmurice, I concede, argued that I was making a silly argument and the calculations I did were moot because the underlying math behind port tunings was an average of port response anyhow. My calculations, however moot, indicated the tuning frequency would theoretically only be altered by a couple hertz anyhow.

The fact remains, this isn't flow in a traditional fluid mechanics sense, but oscillation of a mass of air. The bends may in fact induce minor losses, but to what ultimate effect?

1.) They may induce turbulence and greater friction losses and thus create port noise.
2.) They may alter the tuning frequency.

However, it is still a Helmholtz resonator and is going to act (relatively) as expected and as we want it to. The prudent thing I would think would be to build it with the ability to change the port length slightly after taking an impedance measurement. I don't believe our modeling can accurately predict port bends yet, so empirical measurement with the understanding theory will get you close is the best method.

My opinions of course.
Greg

[Jeff B.]

Re: # of elbows in a port

I used to debate this often on this board.

There is an old thread where I tried to argue that a 90 degree bend would induce minor losses in the port and skew the tuning frequency. Bill Fitzmurice, I concede, argued that I was making a silly argument and the calculations I did were moot because the underlying math behind port tunings was an average of port response anyhow. My calculations, however moot, indicated the tuning frequency would theoretically only be altered by a couple hertz anyhow.

The fact remains, this isn't flow in a traditional fluid mechanics sense, but oscillation of a mass of air. The bends may in fact induce minor losses, but to what ultimate effect?

1.) They may induce turbulence and greater friction losses and thus create port noise.
2.) They may alter the tuning frequency.

However, it is still a Helmholtz resonator and is going to act (relatively) as expected and as we want it to. The prudent thing I would think would be to build it with the ability to change the port length slightly after taking an impedance measurement. I don't believe our modeling can accurately predict port bends yet, so empirical measurement with the understanding theory will get you close is the best method.

My opinions of course.
Greg

I agree with your opinions, and some of them are actually facts. We are not dealing with air flow as much as we are dealing with an oscillation of an air mass acting on the compliance (spring) of the air mass in the box. Yes, the bend does introduce some losses. This can be modeled in the form of a lower Qp value, but hitting the right value accurately will be difficult. Still, you can see the effects of changes in Qp easily enough by modeling a vented system with a Qp around 140, which is equivalent to a typical large, straight port with flared ends, and then reducing it to a Qp of around 60, which would actually be a port with quite a bit of resistive losses, probably more than we are talking about here. You will see that the effect isn't extreme. And yes, it isn't difficult to measure system impedance and derive the port tuning and characteristic Q empirically, which is, as you say, the best way.

Jeff B.

[curt_c]

[Max_Andrews]

Re: # of elbows in a port

I don't think the bends will alter the tuning much, what I'm concerned about is that the resistance to the air induced by the bends will lead to chuffing at a much lower point than the model predicts, so you might as well just use a straight 2" port instead of a 3" port with two bends, since they will chuff and compress at close to the same level.

I did some digging, turns out each 90 degree bend is a 50% loss! So every 90 degree bend you add is the same thing as cutting your port area in half. Hence, I would not recommend bends as they will lead to compression and/or chuffing. I'd like to test this, would be interested to take a subwoofer and try a staight 3" port, a 3" port with one bend, and a 3" port with two bends, and measure at what SPL each of them compresses.