I haven’t done an apple to apple comparison but from experience I’ve got a feeling (or imagination if you like) that smaller port diameter sounds better (may be the midrange information that is lost is less here?). I’m talking about speaker with less than 6” woofer here. And I have preferred a rear-side port, so if noise IS audible that must not be a critical issue I guess.
But hi-end bookshelf speakers use bigger diameter port, hence my question, why not just the smaller one (as long as it’s length is not more than half of the depth)? If turbulence is a real thing, why not using multiple of smaller diameter ports? Ah, yes, multiple ports, I saw that too in hi-end speakers… But may be because they dont have enough space on fron baffle...
But hi-end bookshelf speakers use bigger diameter port, hence my question, why not just the smaller one (as long as it’s length is not more than half of the depth)? If turbulence is a real thing, why not using multiple of smaller diameter ports? Ah, yes, multiple ports, I saw that too in hi-end speakers… But may be because they dont have enough space on fron baffle...
It's really hard to make a case that smaller ports sound better. If the speaker is not playing very low- for example, if there is an active crossover to a sub- then it would be much easier to make such a case.
Turbulence is real, and if you're trying to make real bass then it's a concern.
Multiple smaller ports will act as a single larger port, therefore they need to be longer than a single smaller port. In ideal cases, groups of small ports behave fairly similarly to single large ports.
Do you have any specific way that you think a smaller port is better than a larger one?
Turbulence is real, and if you're trying to make real bass then it's a concern.
Multiple smaller ports will act as a single larger port, therefore they need to be longer than a single smaller port. In ideal cases, groups of small ports behave fairly similarly to single large ports.
Do you have any specific way that you think a smaller port is better than a larger one?
joe carrow said:
Actually..
Multiple smaller ports have a little less turbulence (when considering an equal sd). They can however have more freq. specific turbulence, particularly HIGHER freq. turbulence that leads to "port whine" depending on their shape and grouping.
A larger port however has the advantage with power compression/spl limiting. In other words it has less port air excursion which allows for a louder playback level at lower freq.s.
btw, I think Jay likes the "turbulence"/air pressure - tends to give a more dynamic presentation at lower spl's.
Well, if we take this from the start...
When designing a bass-reflex system, the start is to tune the port to a frequency that suits the design. This is step one and is done with a formula that looks like this:
fp=c/(2*pi) * sqrt (S/ (L*V) )
where c=345 m/s, S is the cross-sectional area of the port, L is the effective length of the port, and V is the volume of the box.
To get this frequency right is far more important than the size of the port itself, and this means that if we are to compare ports with one another, we should do so for the same fp.
From a closer look at the formula, we can see that for a given fp and box volume, the ratio between S and L needs to be fixed. However, we are free to select any S, as long as L is adjusted accordingly to keep fp constant. So here is an extra degree of freedom, that does not necessarily affect the port tuning, fp. A port with a large area S, should also have a large effective length L.
So how does one choose the S and L?
There are mainly two constraints that are necessary to handle.
One is port overloading. For narrow ports, the airspeed in the port becomes high and turbulent. This leads to noise and also affects the frequency response.
The other is pipe resonances. For wide ports, the length of the port has to be long. If the port is made long enough, the pipe resonances (starting at ½ wavelength = the port length) move towards a low enough frequency to interfere with the operation of the speaker.
For large boxes and high fp:s these constraints are unproblematic; it is easy to find port dimensions that satisfy both conditions. On the other hand, for small boxes and low fp:s, the case is the opposite. For such boxes, one may need to use a passive radiator instead of the port.
There are a few things, apart from airspeed that affects the tendency of the airflow to become turbulent, and that is the shape of the cross-sectional area, S (it can be split in several tubes, it can be rectangular etc). Also, flanging the ends of the tube is always good.
To minimize air turbulence for a given area S, the ends should be flanged, and (I am not perfectly sure about this) if S is large, it might be wise to split it into several tubes. That is, it might be better to use two 4" tubes than one 6" tube or make the port rectangular with a rather high aspect ratio.
When designing a bass-reflex system, the start is to tune the port to a frequency that suits the design. This is step one and is done with a formula that looks like this:
fp=c/(2*pi) * sqrt (S/ (L*V) )
where c=345 m/s, S is the cross-sectional area of the port, L is the effective length of the port, and V is the volume of the box.
To get this frequency right is far more important than the size of the port itself, and this means that if we are to compare ports with one another, we should do so for the same fp.
From a closer look at the formula, we can see that for a given fp and box volume, the ratio between S and L needs to be fixed. However, we are free to select any S, as long as L is adjusted accordingly to keep fp constant. So here is an extra degree of freedom, that does not necessarily affect the port tuning, fp. A port with a large area S, should also have a large effective length L.
So how does one choose the S and L?
There are mainly two constraints that are necessary to handle.
One is port overloading. For narrow ports, the airspeed in the port becomes high and turbulent. This leads to noise and also affects the frequency response.
The other is pipe resonances. For wide ports, the length of the port has to be long. If the port is made long enough, the pipe resonances (starting at ½ wavelength = the port length) move towards a low enough frequency to interfere with the operation of the speaker.
For large boxes and high fp:s these constraints are unproblematic; it is easy to find port dimensions that satisfy both conditions. On the other hand, for small boxes and low fp:s, the case is the opposite. For such boxes, one may need to use a passive radiator instead of the port.
There are a few things, apart from airspeed that affects the tendency of the airflow to become turbulent, and that is the shape of the cross-sectional area, S (it can be split in several tubes, it can be rectangular etc). Also, flanging the ends of the tube is always good.
To minimize air turbulence for a given area S, the ends should be flanged, and (I am not perfectly sure about this) if S is large, it might be wise to split it into several tubes. That is, it might be better to use two 4" tubes than one 6" tube or make the port rectangular with a rather high aspect ratio.
Svante said:
One thing to keep in mind is that a drivers T/S change as a function of the voltage drive.
This lmeans to keep the frequency right the port should change depending on how loud you play...
I purposely use very high ratio rectangular ports. This increases port resistance and pushes the character of the design more towards an aperiodic nature which is more tolerant of changing parameters as the drive is changed. (and if it ends up being needed a little foam in the ports can push them to completely aperiodic)
This likely isn't directly related to your question. What is (summarizing what has been said above) is poyential port resonances. The port will resonant at frequencies related to the length & to the diameter of the port. A smaller port will resonata at higher frequencies where there is less energy to exite them. As the port gets smaller, audible chuffing becomes more likey. Radiusing the end of the port really helps out here. And keep in mind that in an ideal world both ends of the port look the same. Taken to the limit that means that the inner end of the port should also terminate on its own flat baffle inside the box.
dave
planet10 said:
Well, sort of...
I mean the best thing to acheive would be a driver thatdoes not change with signal level. A driver that does is inherently nonlinear and should be avoided in any good system.
Of course there will always be some degree of nonlinearity, but if they become so strong that they need to be compensated for by the port design, then they are either too strong or the design is on a really advanced/high performance level. I don't think this is something that should typically be included in a DIY project.
To find the right behaviour of the port to balance the behaviour of the driver, advanced measurement equipment and understanding is needed, and attempting to add such compensations without the right equipment and understanding is likely to make things worse rather than better.
For the average DIYer, finding a linear enough driver is THE way to go, IMO.
On the other hand, disencouraging DIYers to test interesting/odd/advanced things is doomed to fail...
I guess my statement, that there are no physical reason why ports are big, should be more precise. What I meant to say is that - everything else even – ports on small speakers should be small from a engineering point of view. But from the marketing point of view big ports on the front is preferable - fancy.
An explanation to why book shelf speakers with small ports sound better might be that they produce deeper bass, IMO it’s very important to have a good balanced SPL levels over large parts of frequency range. But of course there are a lot of compromises/constrains in port tuning, one are that if a bookshelf speaker is put in a bookshelf the port has to be on the front = noise will be bigger a problem.
An explanation to why book shelf speakers with small ports sound better might be that they produce deeper bass, IMO it’s very important to have a good balanced SPL levels over large parts of frequency range. But of course there are a lot of compromises/constrains in port tuning, one are that if a bookshelf speaker is put in a bookshelf the port has to be on the front = noise will be bigger a problem.
Well, I judge a port size by its length, and some common sense
BTW.... the theory that the port should have a baffle attached on the inside opening makes a lot of sense too
Interesting "port" on Troels latest and highly tweaked design
I wonder if we really know the secrets of a ported design - Scottmoose does some interesting things too in combining BR and horn
http://www.troelsgravesen.dk/C17mk4.htm
BTW.... the theory that the port should have a baffle attached on the inside opening makes a lot of sense too
Interesting "port" on Troels latest and highly tweaked design
I wonder if we really know the secrets of a ported design - Scottmoose does some interesting things too in combining BR and horn
http://www.troelsgravesen.dk/C17mk4.htm
planet10 said:
Well, it is of course a question of sound level and how much of nonlinearity one tolerates. And these things are what we see as distortion in a simple distortion measurement, right?
Anyway, do you think that an average DIYer has the equipment and knowledge to compensate for these effects by designing the port in a particular way, or du you think that most would actually make things worse?
Svante said:
This is an issue that is largely ignored. I 1st ran into the discussion on the bass list years ago when the 1st Adire Shivas were being DuMax tested. It wasn't until a recent comment by Dan Wiggins that i groked it in the form i expressed earlier in the thread.
As to dealing with it... i have always had an aversion to most bass reflex boxes (and we are talking about a lot of speakers over the last 35 years). The way that parameters change is complimentary to a sealed box (ie the parameters change with V in a manner that the LF response of a sealed box remains much the same). This points to designing boxes that are tolerant of a range of parameters. I have been designing this way for a long time, not really understanding that i was -- i just liked the performance better. Aperiodic, quarter-wave designs, sealed all seem to be tolerant.
As examples have a look at the modeled response of the many drivers suggested for Bruce (a BVR, where the large horn shaped ports dominant, or look at what happens if you take the bipole diyREF FR ML-TL and only install one driver (it still works).
dave
Let’s look at a speaker where S=4” and L=10”. What if we made our port this way: Simply drill a 1 cm hole in the 1” rear baffle. Quite a constant S/L right? What kind of audible turbulence can we expect to happen from such a hole? The driver (in box) impedance pattern may not look like double-peak vented impedance, but with such a port the speaker may behave like a sealed speaker, damped but with no significant Fs increase.
Thanks for pointing these out. Now I’m pretty sure will use such (line) port. And radiusing an MDF is easier than radiusing a PVC. And about the internal flat baffle (reminded me with The Froy), this a bit clarifies my curiousity regarding optimum port length (relative to the box dimensions) and position. Not sure about the location at the bottom like some designs tho. I don't feel comfortable with that. I think I prefer double line ports, one on top and one at the bottom. But I will start with one line port, horizontal (and why should it be horizontal!?), in the middle such it will balance the driver.
Really? I can’t feel the logic. Ports look cheap to me. But I can feel it that ports on the side baffle wouldn’t do. Turbulence/noise was the only one I could think of (but I don’t agree so much)
Although I was thinking about equal port tuning frequency situations, you are pointing out something that I have on mind. Deeper bass, balanced SPL. Good, so why not tuning the box below Fs? Find a way to avoid issues. And what is balanced SPL? Can we hear a few dB differences? How about room response, baffle step that cannot be controlled?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.