I get the basics but a debate on another thread regarding Sony SS-S9s has got me wondering about cabinet volume. Why does a driver need a certain cabinet volume and when does it need said volume?
They were discussing 2.5 systems which bring me back to an old question: A sub-woofer and woofer sharing the same air space in a ported cabinet. Some have suggested that all calculations be made on the mean value of the driver's T/S parameters - makes sense. But what happens at high power levels? When the woofer becomes affected by the sub's response to frequency it doesn't 'do'. Effectively offering passive radiation and robbing the port of input ?
They were discussing 2.5 systems which bring me back to an old question: A sub-woofer and woofer sharing the same air space in a ported cabinet. Some have suggested that all calculations be made on the mean value of the driver's T/S parameters - makes sense. But what happens at high power levels? When the woofer becomes affected by the sub's response to frequency it doesn't 'do'. Effectively offering passive radiation and robbing the port of input ?
Too little volume and speaker fights against compressed air inside box.
Too large volume and so too little back pressure inside the box can cause x limit to be met sooner.
Too large volume and so too little back pressure inside the box can cause x limit to be met sooner.
https://en.wikipedia.org/wiki/Helmholtz_resonance
The air behind the cone acts like a spring, changing the resonance of the system allowing acoustic amplification at whatever frequency the box is tuned to.
The air behind the cone acts like a spring, changing the resonance of the system allowing acoustic amplification at whatever frequency the box is tuned to.
Nobody is answering the question. The issue is TWO drivers of different specifications occupying the same air space. A woofer cannot occupy the same air-space as a mid-range driver due to the impact of the woofer's rear wave. Two drivers of equal size but different specs offers a new question.
Unless both drivers have identical or very similar parameters it will be a mess.
I suppose the words "woofer" and "subwoofer" are pure marketing speech in this (sony speaker) case. The so called subwoofer will roll off at the baffle step frequency range.
I do! 🙂
the driver probably doesn't care. 😉
It is the designer that wants to tune the resonant system of driver with enclosure (and eventual additional resonator such as bass reflex) to get the desired response, usually flat or slightly sloped downwards to compensate for room gain.
Your original post has posed 3 questions. The first has been answered. Once understood then it follows that the box would be mistuned for either driver since the box would be sized for neither driver since neither driver has the average T/S parameters. Given the slop in acoustical/mechanical systems, it may be a listenable compromise. To truly find out, one would have to build such a system and test how the bass resonance works out. The easier answer for DIYers is to build two separate enclosures so each woofer has its own acoustic suspension and avoid the issue.
But this seems inefficient. (And take into account the T/S parameters for two identical drivers could be +/- 20%).
Scenario 1: 4 cu ft cab. Sub wants 2.5 ft3, woofer wants 1 ft3. Wouldn't each driver use the excess volume according to it's needs?
Scenario 2: 30hz sine wave. Woofer doesn't respond. Sub salivates: Yay, I got the whole 4 ft3 to myself!
On the one hand it doesn't make for a smooth, flat response but I find the concept interesting.
What is meant by "inefficient"? If you mean by use of materials, depends on the end goal. As accurate and precise a response desired, build materials are a secondary thought.
Scenario 1: Subwoofer is overly constrained due to too small a box. The air cushion is too tight/hard. Won't allow the sub/box system resonate at a low enough frequency to be optimum. Usually will cause a large peak higher in the band. The woofer will be under-constrained, the air cushion being to soft/loose. This may allow the woofer to bottom out earlier than expected when the volume is cranked.
Scenario 2: The woofer will respond unless a high-pass filter has been implemented.
Since you find it interesting, build and test to see what it does. the joys of DIY. Speaking of salivating, I'm looking forward to getting this latest move out of the way so I can get back to experimenting.
Scenario 1: Subwoofer is overly constrained due to too small a box. The air cushion is too tight/hard. Won't allow the sub/box system resonate at a low enough frequency to be optimum. Usually will cause a large peak higher in the band. The woofer will be under-constrained, the air cushion being to soft/loose. This may allow the woofer to bottom out earlier than expected when the volume is cranked.
Scenario 2: The woofer will respond unless a high-pass filter has been implemented.
Since you find it interesting, build and test to see what it does. the joys of DIY. Speaking of salivating, I'm looking forward to getting this latest move out of the way so I can get back to experimenting.
With a driver is a separate box the air in the box acts like a spring and the resulting force the spring applied to the driver is dependent only on the motion of the driver. When 2 drivers share the same box the spring is connected to both drivers so the motion of either driver causes a force to be applied to the other driver. So any signal applied to one driver will result is a spring force being applied to the other driver. It is what is called a coupled system where as a diver in its own box is an isolated system.
The general rule is DON’T.
It can be made to work in some few cases, but not often.
I avoid 2 different drivers covering the same range (except maybe in an omni).
dave
Re: Two dissimilar drivers in the same space & 'calculations be made on the mean value of the driver's T/S parameters'
- that over simplifies it a bit, use the Means of Fs & Q values, all other TSPs add. It works well on drivers that aren't too dissimilar
- that over simplifies it a bit, use the Means of Fs & Q values, all other TSPs add. It works well on drivers that aren't too dissimilar
What Planet10 said, "Don't". An enclosed space has acoustic resonances with frequencies dependent on the internal dimensions. Damping materials absorb hid and high frequencies well and are ineffective at absorbing low frequencies. So there is an advantage to having small cabinets for each driver that have relatively high frequency first resonant mode frequency as it will be possible to damp it by filling the space with dense damping material virtually eliminating "box sounds". By experimentation with a low distortion Seas 8" woofer I found the distortion would begin to rise as the box volume was decreased. So having two drivers share a large volume is really bad as it produces a cabinet with low resonances that will be poorly damped. Ported speakers in general produce low fidelity bass as during the first cycle of any exponential decaying plucked bass note, so most of them, near the port frequency gets crushed as the only output is from the woofer until the port starts to ring after a few cycles. Well by then the exponential decay has reduced the amplitude. So they destroy bass dynamic impact.
My personal opinion is that even pretty small enclosures are tolerable and can be EQed using Linkwitz approach to it. As long as you use a passive radiator.
Played loud compressed bass music for hours with Behringer inuke Nu3000dsp but the speaker never showed thermal problems. It was like +11db at 40hertz but - 6.5 db at 150hertz, 20cm driver in 8 Liter box
Here some example with 18inch drivers.
https://www.diyaudio.com/community/...how-off-my-new-18in-build.365764/post-6767025
Played loud compressed bass music for hours with Behringer inuke Nu3000dsp but the speaker never showed thermal problems. It was like +11db at 40hertz but - 6.5 db at 150hertz, 20cm driver in 8 Liter box
Here some example with 18inch drivers.
https://www.diyaudio.com/community/...how-off-my-new-18in-build.365764/post-6767025
BUT
Many older Japanese speakers did use 2 dis-similar drivers sharing the same box volume as did several well regarded speakers from Wharfedale; so it can and has been done before and we do it all the time with OB speakers and if we consider an OB to be an infinitely large box I see no problem, so long as the drivers are not over powered.
I think it works if the drivers are sharing the same frequency band.
Say a 15 and a 10 working up to where a smaller driver carries the upper midrange and treble, that small driver needs it's own enclosure but in early ones such drivers were an integral sealed back unit.
I'm probably wrong but my limited understanding is that the port, if used, was tuning the box not the driver.
Many older Japanese speakers did use 2 dis-similar drivers sharing the same box volume as did several well regarded speakers from Wharfedale; so it can and has been done before and we do it all the time with OB speakers and if we consider an OB to be an infinitely large box I see no problem, so long as the drivers are not over powered.
I think it works if the drivers are sharing the same frequency band.
Say a 15 and a 10 working up to where a smaller driver carries the upper midrange and treble, that small driver needs it's own enclosure but in early ones such drivers were an integral sealed back unit.
I'm probably wrong but my limited understanding is that the port, if used, was tuning the box not the driver.
You tune the box to the driver. It is a musical instrument of sorts. You're using the leaky (ported) or non-leaky (sealed) mass of air to enhance the performance of the driver by changing where that driver resonates.
OB is a different beast because the air-mass is so loose, it doesn't contribute to the performance of the drivers. Many folks look for a driver Qts of ~0.6 as it provides a "nicely" damped response.
Sure, mixing dissimilar woofers can be done. The question becomes whether it leads to optimal or sub-optimal performance and what the end system goal is. I'm all the "have at it and see what it does" type. Stop worrying about what the theorycrafting says. Experiment and build intuition backed by theory.
OB is a different beast because the air-mass is so loose, it doesn't contribute to the performance of the drivers. Many folks look for a driver Qts of ~0.6 as it provides a "nicely" damped response.
Sure, mixing dissimilar woofers can be done. The question becomes whether it leads to optimal or sub-optimal performance and what the end system goal is. I'm all the "have at it and see what it does" type. Stop worrying about what the theorycrafting says. Experiment and build intuition backed by theory.
Last edited:
hell. pull a Thiele & Small. Buy a bunch of drivers, build a bunch of different boxes and start measuring to see what the correlations are to characterize how to formulate the T/S parameters for boxes with dissimilar woofers. Write a paper and get it published in the acoustic science journal thingy.
I say: do!
You actually can simulate two drivers in one enclosure, but you have to write a few lines of code yourself as no app provides the option.
Quite a few modern DIY designers went this way, Troels Gravesen and Tony Gee IIRC. Ralph Smulders once used a Briggs filter for one of two identical drivers in a very popular design then.
To me, there seems no real advantage. The enclosed air has to behave as a spring in the whole resonant range of a driver, about 1,5 to 2 octaves around fc. In most designs, the ranges will overlap, so every driver would need its own amount of volume anyway.
You actually can simulate two drivers in one enclosure, but you have to write a few lines of code yourself as no app provides the option.
Quite a few modern DIY designers went this way, Troels Gravesen and Tony Gee IIRC. Ralph Smulders once used a Briggs filter for one of two identical drivers in a very popular design then.
To me, there seems no real advantage. The enclosed air has to behave as a spring in the whole resonant range of a driver, about 1,5 to 2 octaves around fc. In most designs, the ranges will overlap, so every driver would need its own amount of volume anyway.