Speaker Enclosure

[floppygoat]

What do enclosures actually do? How do they work? How do the mechanical/electrical parameters correspond with enclosure size and type? Etc...

Thanks,

Tony

[billfitzmaurice]

http://techtalk.parts-express.com/showthread.php?t=219617

[diy speaker guy]

This is a massive topic, but to get started:

An enclosure prevents the front of the cone from directly interfering with the back, which would result in cancellation below the baffle step frequency. So with a box, even a sealed box, you get more bass than running the driver naked, or even naked except for a baffle.

All boxes (even small sealed boxes) create a series of impedance peaks that correspond to their internal physical dimensions. In sealed boxes, the boxes are usually stuffed (when necessary to obtain the desired system q) and the stuffing mostly or totally kills all the box resonances. Even with no stuffing the internal resonances shouldn't be too significant to the overall response, but if the box size starts to get large wrt the wavelengths being played, the resonances should at least be considered.

Boxes with an opening are generally meant to use the resonances created within the box, at least the lowest fundamental resonance. This serves to provide an acoustical bass boost at frequencies that correspond to these resonances.

The parameters that are responsible for box size are qts, vas and fs. High q and high vas = big box. Drivers with a q of .4 will usually work well in a box size = vas. Higher q needs larger, lower q needs smaller.

Low q drivers (less than .4) are generally used in horns, since it takes a bunch of impedance peaks to prop up the low frequency response. Moderate q drivers (.3 - .6 or so) are generally used in ported boxes (or mltl, tl, or other 1/4 wave box). Higher q drivers (.5 - .9) are usually used in sealed boxes, and high q drivers (.9 or more) are usually only good for OB. .7 is considered just about perfect for IB but if you have eq it doesn't really matter much.

This is all very very general but it's a start.

[floppygoat]

As with mids and high, does the "sound" emanate from the cone of a subwoofer? If so, then why is greater sound emanated from a ported enclosure when the cone movement is minimized, almost stopped, at tuning. Also, then why are subs usually orientated backwards, as if the sound was emanating from the box, not the cone of the sub? I have heard the the box is used to "enclose" the rear emanations to prevent cancellation. However, this is counter intuitive for it being the only purpose.

[floppygoat]

http://techtalk.parts-express.com/showthread.php?t=219617

Could you direct me to a more specific link?

Thanks though.

[diy speaker guy]

As with mids and high, does the "sound" emanate from the cone of a subwoofer? If so, then why is greater sound emanated from a ported enclosure when the cone movement is minimized, almost stopped, at tuning. Also, then why are subs usually orientated backwards, as if the sound was emanating from the box, not the cone of the sub? I have heard the the box is used to "enclose" the rear emanations to prevent cancellation. However, this is counter intuitive for it being the only purpose.

Yes, the sound comes from the cone, but if there is a port, the port is another additional source of sound. The port (very generally speaking) is only supposed to emit a narrow band of bass frequencies. The cone doesn't have to move far at the resonant frequency because the box/port system provides the acoustical boost at that frequency. It doesn't matter much which way you face a sub, since the wavelengths in question are so large.

[diy speaker guy]

There are a few different ways to learn this stuff. I imagine Ray Alden's book is perfect for this, or you can just google your way through sealed, helmholtz and 1/4 wave theory. Or you can see all this in effect by using and playing around with WinISD, or even better Hornresp.

[floppygoat]

Or you can see all this in effect by using and playing around with WinISD, or even better Hornresp.

I am very familiar with win and how to design enclosures. I am also very aware of how different enclosures effect different speakers with different mechanical and electrical values. I am also aware of the different enclosure and how they work. I am interested and why, and what/how the actual physical properties work to create these changes.

[floppygoat]

Yes, the sound comes from the cone, but if there is a port, the port is another additional source of sound. The port (very generally speaking) is only supposed to emit a narrow band of bass frequencies. The cone doesn't have to move far at the resonant frequency because the box/port system provides the acoustical boost at that frequency. It doesn't matter much which way you face a sub, since the wavelengths in question are so large.

This really doesn't make sense to me. Since the output can see an ideal 3db boost at tuning, the displacement should theoretically double (at least according to my little mind). Therefore, the volume of the air coming out of the port would need to be equivalent to the difference of ((2 x air displaced by cone at tuning frequency in sealed enclosure) - cone displacement at tuning in a bass reflex). Since the port area is usually much smaller then cone area, I find this very unlikely. Maybe this is actually the case. If so, then it makes sense...

[diy speaker guy]

I am very familiar with win and how to design enclosures. I am also very aware of how different enclosures effect different speakers with different mechanical and electrical values. I am also aware of the different enclosure and how they work. I am interested and why, and what/how the actual physical properties work to create these changes.

Are you looking at ALL the screens in WinISD? Check the impedance, and port spl of a ported box and you will see a single large, narrow spike. That's the helmholtz resonance of the box. You can think of it as an acoustical bass boost, filling in below where the driver's cone response is falling off.

Resonances can be very powerful, the larger the box is the more acoustical boost you can get. In WinISD make the box larger and smaller and watch the impedance peak and port spl change.

Then, read MJK's alignment tables article. http://www.quarter-wave.com/TLs/Alignment_Tables.pdf This will explain how geometry (shape), stuffing, and driver parameters affect things. This article is about transmission lines but the concepts are generally the same for all enclosure types.

[diy speaker guy]

This really doesn't make sense to me. Since the output can see an ideal 3db boost at tuning, the displacement should theoretically double (at least according to my little mind). Therefore, the volume of the air coming out of the port would need to be equivalent to the difference of ((2 x air displaced by cone at tuning frequency in sealed enclosure) - cone displacement at tuning in a bass reflex). Since the port area is usually much smaller then cone area, I find this very unlikely. Maybe this is actually the case. If so, then it makes sense...

Displacement doesn't need to double if you have a strong resonance to back it up. It's like banging a stick on concrete. The sound will be a short, highish frequency snap. Now bang the stick on a very large empty metal fuel holding tank and you will hear a lower frequency prolonged bong. You didn't change the way you hit with the stick but yet it sounds different. That's resonance at work. That's what a resonant enclosure does, basically.

[killa]

The resonance gets louder the closer you get to tuning. As it gets louder insise the box it tries to escape and pushes the cone (poor wording) Impeding it's movement. This is where the impedance spike comes from. At least thats how I picture it in my mind. Please correct if wrong

[billfitzmaurice]

Since the port area is usually much smaller then cone area, I find this very unlikely.Google 'Bernoulli's Principle'. It explains why port area does not have to equal cone area, and why chuffing occurs when port area is insufficient.

[floppygoat]

Thanks guys. I will research it some more, contemplate what I have been told, and get back with questions if I have any.

Tony

[floppygoat]

Thanks so much diy speaker guy, I re-read your comments after some research and it makes more and more sense, and is more and more helpful each time I read it. I am still working my way through it, finding I just run into more questions. My personalty of "needing" a reason why is sometimes frustrating, i know... :o

[floppygoat]

All boxes (even small sealed boxes) create a series of impedance peaks that correspond to their internal physical dimensions. In sealed boxes, the boxes are usually stuffed (when necessary to obtain the desired system q) and the stuffing mostly or totally kills all the box resonances. Even with no stuffing the internal resonances shouldn't be too significant to the overall response, but if the box size starts to get large wrt the wavelengths being played, the resonances should at least be considered.


What is meant by "impedance." Tried looking it up, and don't understand the "acoustical" and "mechanical" definitions.

Also, a speaker has a natural resonance. A sealed box does, a ported box does, everything does...correct? When the speaker/sub are placed in a box, these are combined to create a combined resonance (system q). Now looking at a sealed box system, as I gathered from what you said, the box resonance has little affect on the system, actually, from my experience raises the resonance of the system. Why below this system q does the output begin to decline at 12db/octave? Even in a infinite baffle design, where speaker performance is completely unassisted/unhindered, is there still such a roll off? Does this roll off hold true with any speaker below resonance (tweeters, mids, woofers, subs)?

[rone]

What is meant by "impedance." Tried looking it up, and don't understand the "acoustical" and "mechanical" definitions.

In electrical systems Z=V/I (Volts / current - vector quantities)
Impedance is the volts across a device divided by the current that flows through it.

In mechanical systems Z=F/U (force/velocity)
In acoustical systems Z=p/U (pressure/velocity)
Impedance is the pressure or force between the two terminals of the device divided by the velocity that is created by that force. In a lower impedance more velocity is created for a given force and vice versa.

Things get complicated quick from here, as there is an inverse analogy called mobility that is = 1/Z that is often more convenient.

A sealed box rolls off at 12dB/octave because it has essentially constant excursion below resonance and the output of a constant excursion device falls at 12dB/octave. This is true for any moving coil transducer including tweeters and mids.

[diy speaker guy]

What is meant by "impedance." Tried looking it up, and don't understand the "acoustical" and "mechanical" definitions.

Also, a speaker has a natural resonance. A sealed box does, a ported box does, everything does...correct? When the speaker/sub are placed in a box, these are combined to create a combined resonance (system q). Now looking at a sealed box system, as I gathered from what you said, the box resonance has little affect on the system, actually, from my experience raises the resonance of the system. Why below this system q does the output begin to decline at 12db/octave? Even in a infinite baffle design, where speaker performance is completely unassisted/unhindered, is there still such a roll off? Does this roll off hold true with any speaker below resonance (tweeters, mids, woofers, subs)?

Yes, everything has a resonant frequency. Resonances are everywhere, there's more of them than you might think. For example, a driver has a resonant frequency (fs) but the cone itself also sometimes has it's own resonances (energy storage issues, aka breakup). Both the fs and other cone resonances will show up in an impedance sweep as a spike in resistance at those frequencies. The bigger the impedance peak the stronger the resonance.

Enclosures have a series of internal resonances, even the smallest sealed boxes. Ports have their own resonances. Simple simulation programs grossly simplify resonances and express them as a single fundamental resonance (helmholtz for example) instead of a series of impedance peaks that 1/4 wave simulation programs more thoroughly identify.

Resonances are described by quality factor, or q. High q resonances are sharp narrow spikes, low q resonances are broadband but lower in amplitude. The first picture on this page http://sites.google.com/site/amateuraudio/theory/baffle-effects shows different values of driver qts. As you can see, the rolloff below resonance depends on the driver q. The same is true for systems (drivers in enclosures). Below the fundamental resonance the rolloff is largely determined by the system q.

[floppygoat]

As you can see, the rolloff below resonance depends on the driver q. The same is true for systems (drivers in enclosures). Below the fundamental resonance the rolloff is largely determined by the system q.

So it is not necessarily a 12db/octave roll off? How is impedance related to resonance?

[floppygoat]

A sealed box rolls off at 12dB/octave because it has essentially constant excursion below resonance and the output of a constant excursion device falls at 12dB/octave. This is true for any moving coil transducer including tweeters and mids.

Thanks, this makes sense. Since the driver is most "efficient" at resonance, then excursion is maximized. Driving hzs below this frequency will require more excursion, thus greater power. However, power is constant, so excursion cannot increase. This results in a loss in output. Only alternative is to boost power, or port. The port resonance then couples the drivers and acts as a boost to the system. Correct?

[floppygoat]

Also, like with cabin gain in a room or car, wouldn't the the rear-ward waves of the speaker create a similar boost within the enclosure? Would these be audible and/or add to the output? Maybe this is the point of the enclosure...hell if I know, just trying to work through the sciences behind it all.

Now for ports, is it the air itself that is creating a resonance when in moves in and out, or it is the port itself? If it were the port, makes me think this would be very dependent of the port material. Since emphasis on the port material is rarely placed, I am assuming its the air. I understand the principle of porting, but in case I don't, I will describe how I believe it works. As the driver pushes air into the enclosure at a frequency above tuning, the air within is compressed. This forces air out of the port. However, the driver quickly returns to neutral and then out, sucking air back into the port. The closer you get to tuning, the larger the volume of air will become which is being forced in and out of the port. When the volume of the air being compressed and expanded is equivalent to that of the port, the ports resonance is reached.
Correct?
If so, this should mean that even below tuning, the port is resonating but the speaker is not. This is because, like free air, the rear waves begin cancelling.
Correct?

[diy speaker guy]

So it is not necessarily a 12db/octave roll off? How is impedance related to resonance?

Sealed boxes will want to roll off at 12 db/oct but depending on the size of the box, you may be able to change the low knee and lower frequencies on the curve to something other than 12 db/oct, at least for awhile until they normalize. Different alignments will have different rates of roll off as well, I think ported rolls of at 18 db/oct below tuning. And again, you can manipulate that somewhat by changing your frequency response curve.

How is impedance related to resonance? An enclosure resonance will show up in an impedance sweep as a spike. This corresponds to a reduction in excursion for the frequency range involved.

[diy speaker guy]

Also, like with cabin gain in a room or car, wouldn't the the rear-ward waves of the speaker create a similar boost within the enclosure? Would these be audible and/or add to the output? Maybe this is the point of the enclosure...hell if I know, just trying to work through the sciences behind it all.

Yeah, "cabin gain" is based on size and shape of the cabin(et). The size determines how much acoustical gain potential there is (the strength of the resonance(s)), and the shape of the cabin (enclosure) determines the location(s) of the impedance peak(s). That's all happening inside the box. Then we can use a port if desired, to tune the system to a desired frequency and let the energy inside the box out to sum with the direct radiation from the driver cone.


Now for ports, is it the air itself that is creating a resonance when in moves in and out, or it is the port itself? If it were the port, makes me think this would be very dependent of the port material. Since emphasis on the port material is rarely placed, I am assuming its the air. I understand the principle of porting, but in case I don't, I will describe how I believe it works. As the driver pushes air into the enclosure at a frequency above tuning, the air within is compressed. This forces air out of the port. However, the driver quickly returns to neutral and then out, sucking air back into the port. The closer you get to tuning, the larger the volume of air will become which is being forced in and out of the port. When the volume of the air being compressed and expanded is equivalent to that of the port, the ports resonance is reached.
Correct?
If so, this should mean that even below tuning, the port is resonating but the speaker is not. This is because, like free air, the rear waves begin cancelling.
Correct?

No, not really. There's loads of information on how ports work, with animations and stuff. A mass on a spring. There's a lot of math that defines these relationships and if you haven't understood the classical definitions I'm not sure how much I can help. But like I recommended before, playing with a simulation program and flipping through all the windows (excursion, port output, cone output, impedance, frequency response, etc) and making changes and re-examining those windows can show you in pictures what the written theory is not helping you with.