I would like to know your opinion about the pros and cons of the location of port resonance in a bass reflex speaker. The assumption is that a vented speaker can be build with n ports to achieve the same tuning frequency fb. The low limit for the number of ports is given by one's own decision where the threshold of airspeed should reside, which is also dependent on listening levels. Often, a single port is okay, sometimes two can be desireable. All this said, one can use port numbers to deliberately choose a frequency range wherein the unavoidable port resonance will occur.
To give a practical example, I show options from a speaker I am working on at the moment. The base formula to calculate the center of the first port resonance F1 is:
Wherein c = speed of sound (m/s), L = vent length (m) and d = vend diameter (m)
I only want focus on length here, because this is the easiest way to change the properties of a commercial vent, continuing the example: For a 50 liter box and a box tuning of fb = 35 Hz and a vent diameter of 8 centimeters/surface of 50 square centimeters, with one port, vent length equals 18.7 centimeters, while with two ports, vent length is 43.3 centimeters.
F1a = 343/(2*0.187+0.8*0.08) = 785 Hz
F1b = 343/(2*0.433+0.8*0.08) = 368 Hz*
If I am correct with my above assumptions, I may by changing the number of ports choose, which area of the response will by affected by port resonance. Is there some more insight into this written down? Port resonance does introduce cancellations and, I don't really know, maybe also distortion. Therefore I would assume it is important to think about where this is most audible. I would be happy if someone can show me further literature or give his*her own experiences.
Best
m.
________________
* I am not sure if the lengths add up with two vents and port resonance with mutiple vents has to be calculated as and if it was one total system. Further, neither do I know if only length or also diameter must be added. For such a case F1c = 343/(2*0.866+0.8*0.16) = 184 Hz, which I think is unrealistically low, which is why I rejected this calculation.
To give a practical example, I show options from a speaker I am working on at the moment. The base formula to calculate the center of the first port resonance F1 is:
Code:
F1 = c/(2*L+0.8*d)I only want focus on length here, because this is the easiest way to change the properties of a commercial vent, continuing the example: For a 50 liter box and a box tuning of fb = 35 Hz and a vent diameter of 8 centimeters/surface of 50 square centimeters, with one port, vent length equals 18.7 centimeters, while with two ports, vent length is 43.3 centimeters.
F1a = 343/(2*0.187+0.8*0.08) = 785 Hz
F1b = 343/(2*0.433+0.8*0.08) = 368 Hz*
If I am correct with my above assumptions, I may by changing the number of ports choose, which area of the response will by affected by port resonance. Is there some more insight into this written down? Port resonance does introduce cancellations and, I don't really know, maybe also distortion. Therefore I would assume it is important to think about where this is most audible. I would be happy if someone can show me further literature or give his*her own experiences.
Best
m.
________________
* I am not sure if the lengths add up with two vents and port resonance with mutiple vents has to be calculated as and if it was one total system. Further, neither do I know if only length or also diameter must be added. For such a case F1c = 343/(2*0.866+0.8*0.16) = 184 Hz, which I think is unrealistically low, which is why I rejected this calculation.
Well, a duct is a 1/2 wave resonator so has a fundamental (end-correction notwithstanding) at lambda/2. By and large, you really don't want duct resonance at all; it tends to become more problematic for obvious reasons as length increases, shifting the fundamental lower in frequency. As a general ROT my own limit is a duct length of 6in, especially if it's forward-facing. Rear or down-firing where boundary etc. losses may come into play can be a little more forgiving. If length needs to significantly exceed that, then a slightly smaller vent CSA may be preferable, accepting a minor increase in Vmax through the duct (again, often compensated for by rear or down-firing), or if it is very extreme, shifting to passive radiators, which in essence are most applicable for drive units that can be tuned low in an acoustically small volume.
Hello Scottmoose,
thank you for sharing your experiences. Can you give some more detailed description of what you mean with the following, also adding in some variables, like what kind of box this relates to?
added: My general understanding has always been that resolution of hearing is smaller in the lower frequencies and this may contradict the placement of the port resonance at a higher frequency.
thank you for sharing your experiences. Can you give some more detailed description of what you mean with the following, also adding in some variables, like what kind of box this relates to?
why is this more problematic?
added: My general understanding has always been that resolution of hearing is smaller in the lower frequencies and this may contradict the placement of the port resonance at a higher frequency.
But they will always be there, because the laws of nature. Or did I miss something? Dampening will only reduce them.
Last edited:
Not 'experiences' per se aside from a personal design ROT -primarily just physics. 😉 Vented boxes are all just variations on a theme as far as this is concerned; however, Onkens are one of the types (please note I do not use 'one of' as a synonym for 'only') most susceptible to problematic self-harmonics due to their large & typically lengthy duct systems. That may in certain cases be why some like them, which is fair enough, but not something I would advocate.
The lower in frequency the vent fundamental is, the greater the likelihood tends to become of it being audibly excited.
Yes, I'm well aware of that. The point I was making is that in an ideal world, you don't wish for there to be any duct resonance, ergo the objective consonant with good overall design is to achieve a practical minimum without excess sacrifice to other qualities.
The lower in frequency the vent fundamental is, the greater the likelihood tends to become of it being audibly excited.
Yes, I'm well aware of that. The point I was making is that in an ideal world, you don't wish for there to be any duct resonance, ergo the objective consonant with good overall design is to achieve a practical minimum without excess sacrifice to other qualities.
Last edited:
Using two ports is primarily the same as using one wider port. We can understand what you mean, but I hope that you're meaning width (combined) and length of port, and not thinking that it has to do with how many there are.
Hello Allen,
I am not really sure what you are relating to, so I will freely apply your information to the formula for port resonance which I took from a techtalk thread. As it predicted port resonance location quite well in my case, I assumed it is right, but if there is any better, I would take a hint.
Now my current understanding of port resonance is that it occurs primarily as a function of port length, whilst port surface does add some small amount of modulation too. Therefore, port length primarily defines the frequency in which resonance occurs. If I increase surface, for a given resonance frequency fb, port length must be increased. Here I understand you that adding a second port is essentially the same as increasing surface and both imply a longer pipe and therefore a different position of resonance within the speaker's output. Strictly speaking, it is not a matter of the count of ports, just that if one does use commercial ports, it is easier to use a second identical port instead of making it wider.
You wrote: width (combined) and length, this reads as if width is adding up but length is not, as in:
F1p1p2 = 343/(2*L((p1+p2/2)+0,8*d(p1+p2))
Is this right?
I am not really sure what you are relating to, so I will freely apply your information to the formula for port resonance which I took from a techtalk thread. As it predicted port resonance location quite well in my case, I assumed it is right, but if there is any better, I would take a hint.
Now my current understanding of port resonance is that it occurs primarily as a function of port length, whilst port surface does add some small amount of modulation too. Therefore, port length primarily defines the frequency in which resonance occurs. If I increase surface, for a given resonance frequency fb, port length must be increased. Here I understand you that adding a second port is essentially the same as increasing surface and both imply a longer pipe and therefore a different position of resonance within the speaker's output. Strictly speaking, it is not a matter of the count of ports, just that if one does use commercial ports, it is easier to use a second identical port instead of making it wider.
You wrote: width (combined) and length, this reads as if width is adding up but length is not, as in:
F1p1p2 = 343/(2*L((p1+p2/2)+0,8*d(p1+p2))
Is this right?
Last edited:
more ports will give a bigger overall port mouth area, and consequently longer ports, which increases the chances of internal duct resonances. For domestic listening, i find that a single port of ~50mm dia is adequate, and short enough that damping can be placed on the wall behind it so that mid reflections are minimised
Why is that so? Here is an example of the port resonance of a box without dampening. The resonance is peaking at 600 Hz, because the vent is 25 cm long 8 cm diameter, but it also stretches 100 to 300 Hz wide, depending on the pressure level one finds to be of relevance. Is a port resonance only excited if within the source material, the precise frequency (let's say +-5 Hz) is played, or also in a wider range? Is there a function with something like a Q value to define the steepness of the sides of a resonance frequency, where one can tell within which range it is excited?
Attachments
Last edited:
You just posted this (correct information) at an inopportune time due to the question I was asked 😉 So to clarify...
Adding a second port where there used to be one, increases the combined width but does not add up to a longer port.
However as you note it changes the tuning and so creates the need to then make it longer.
Hello Pete,
to my understanding, port resonance is not something that occasionally occurs, but is unavoidably a property of this sort of resonator device. It can only be suppressed by dampening to a level which is effectively masked by the drivers output. In bad cases, a port resonance might create a single, very audible frequency which can be clearly heard under a given listening situation, depending on levels, source material and so on.Then, there is "chance" involved, but only insofar as the rest of the resonance is effectively masked or does not show this behavior, while still a constant property and observable in a measurement of port output.
I understand why the length of the port must be increased if one wants to keep fb the same. My question is about the right application of this information with the given formula. Is the correct use for Length = (length port 1 + length port 2/2), which assumes that resonance is a function of the single physical duct device, despite that fb is a result of the virtual total port area which is usually a sum of port diameter = (diameter port 1 + diameter port 2). Or is length of the vents also added to one another, as in Length = (length port 1 + length port 2)?
To return to my example:
As is obvious I did not use a sum of port area with the second formula. It must be F1b = 343/(2*0.433+0.8*0.16) = 345 Hz. Here, port area is a sum of both physical devices, while length is the length of one physical device.
Last edited:
Its not just port resonance..... box standing wave also leaking through the port opening. It looks like you have port resonance near box standing wave, so you geting wide response.
Box standing wave you can reduce with damping material, for port resonance you can make trap by driling the holes at half length of port and then cover that holes with material which doesnot leaking air, but flexing. Neoprene foam working.
Last edited:
Hello Radule,
I saw a patent on google patents today which explains this, but did not think this was necessary. I assume damping will solve a lot, my question is actually much more about whether port resonance frequency can be optimized by choosing a frequency location which is better and what criteria applies. As in my example the choice between 345 or 785 Hz.
If it is bass driver in 3 Way, then obviously is better to have port resonance at 785hz, assuming xover is lower in frequency, so resonace will be less excited.
For 2 way, i can not to say much.
For 2 way, i can not to say much.
Hello Allen,
thank you for clarifying that.
The correct formula for two ports in my example then is:
F1 = 343/(2*0.433+0.8*0.16) = 345 Hz
or generalized
F1 = 343/(2*L+0,8*d)
Wherein L = port length of n ports/n (single port device length)
d = sum of port area of n ports (sum of all port devices
Easiest way is to try and get the 1st port resonance above the passband of the woofer. A starting point is to model the driver at xmax and get the port velocity ~17ms, then see what frequency the 1st resonance is at.
If it is still in the passband maybe consider a 3 way.
If it is still in the passband maybe consider a 3 way.
Apart from that I can neither change to a 3-way topology nor that I see a possibility to move the resonance high enough beyond the crossover, out of curiosity for my education:
By the way, until now I have not heard any resonance, but then again I am not a trained listener who knows what sounds right and what not in audio engineering. My main interest was whether it can be used as a tool in speaker design to choose a higher or a lower first port resonance frequency to optimize acoustic reproduction. Until now, I could hear the opinion that higher is better, which contradicts my understanding of aural and psychoacoustic resolution, but which makes sense in the way that higher fundamental can shift distorted harmonics. Is there other opinions around?
Isn't a resonance a property of the mechanical system and even if the crossover cancels woofer electrical energy at port resonance frequency, it still gets excited because of the frequencies below that which build up the resonance? And which, accordingly, even will take place at F1 when no electrical impulse drives the the motor at this given frequency?
By the way, until now I have not heard any resonance, but then again I am not a trained listener who knows what sounds right and what not in audio engineering. My main interest was whether it can be used as a tool in speaker design to choose a higher or a lower first port resonance frequency to optimize acoustic reproduction. Until now, I could hear the opinion that higher is better, which contradicts my understanding of aural and psychoacoustic resolution, but which makes sense in the way that higher fundamental can shift distorted harmonics. Is there other opinions around?
The frequencies below the port resonance might not excite it by themselves but their harmonics which are caused be the woofer's nonlinearities might do so.
As already mentioned there are tricks that minimise this resonance. And by how much it is excited depends on the location of the port and the internal damping of the cabinet as well.
Maybe you can high-pass the signal going to your speaker such that the low-end influence of the reflex port is minimised and then you listen whether you can hear differences in midrange reproduction between open and closed reflex port.
Regards
Charles
As already mentioned there are tricks that minimise this resonance. And by how much it is excited depends on the location of the port and the internal damping of the cabinet as well.
Maybe you can high-pass the signal going to your speaker such that the low-end influence of the reflex port is minimised and then you listen whether you can hear differences in midrange reproduction between open and closed reflex port.
Regards
Charles