Iv'e been looking at the Dayton Audio RSS265HF-4 10" Reference HF Subwoofer 4 Ohm and getting somewhat confused about the optimal enclosure size. The manufacturer recommends a 0.53 cubic ft sealed enclosure but I can see in the review section on the Parts Express site that people have been making all sorts of different size enclosures https://www.parts-express.com/dayto...erence-hf-subwoofer-4-ohm--295-460#lblReviews
Does this mean that there is a huge margin of error or is there several 'optimal' size enclosures for each driver?
Thanks!
Myles
Does this mean that there is a huge margin of error or is there several 'optimal' size enclosures for each driver?
Thanks!
Myles
The enclosure will determine final performance. So, yes, they'll all be different. For a sealed enclosure, small differences in box size have little effect. Also, stuffing density can increase apparent volume.
The question is: Optimal for what application?
The biggest contrast I can think of off the cuff is the difference in requirements between a sub for car use and that for home use.
I recommend you simulate the various enclosures using a tool like Unibox or WinISD. Check what happens to frequency response, group delay and excursion.
The question is: Optimal for what application?
The biggest contrast I can think of off the cuff is the difference in requirements between a sub for car use and that for home use.
I recommend you simulate the various enclosures using a tool like Unibox or WinISD. Check what happens to frequency response, group delay and excursion.
As you go smaller than optimal, bass drops off fast and you start get a hump above the roll off. Some oldschool car audio subs needed HUGE boxes to be "optimal" in a normal listening environment. But in a smaller enclosure, they increase a little in output due to hump, and deep bass is augmented by transfer function of vehicle. And a little more powerhandling, too. Just a little trick.
With winisd pro, you can see how power handling is effected.
Experiment with stuffing, too.
I have subs that simmed optimally sealed at 5.5 cu ft, and manufacturer recommended 1.75. They sound awesome in a car. Pretty good in the house, too.
Linkwitz transform is one way to go small enclosure if you have power handling, xmax and power.
With winisd pro, you can see how power handling is effected.
Experiment with stuffing, too.
I have subs that simmed optimally sealed at 5.5 cu ft, and manufacturer recommended 1.75. They sound awesome in a car. Pretty good in the house, too.
Linkwitz transform is one way to go small enclosure if you have power handling, xmax and power.
A useful range of Qtc is ~0.5 to 1.1 or so.
Qtc=Qts*sqrt(Vas/Vb+1) or Vb=Vas/((Qtc/Qts)^2-1)
For a sealed subwoofer, a Qtc of 0.707 is a so-called "maximally flat" design. For that value of Qtc and the published specs, a box of about 1.3 cubic feet is necessary. Using some damping material in the box will allow reducing the volume to a cubic foot or so.
The recommended box would give a Qtc of about 1.
Higher Qtc than 0.7 gives a bit of peaking, lower than 0.7 gives some drooping.
Only one parameter: size. The bigger, the less you are crippling the resonance frequency. "Critical damping" is a slogan, just a cute term borrowed from electrical analogies, not an acoustical goal.
The box is put there to solve the wave cancellation riddle, not because it does anything helpful. Infinitely big would be best. Anybody disagree?
B.
The box is put there to solve the wave cancellation riddle, not because it does anything helpful. Infinitely big would be best. Anybody disagree?
B.
Thanks so much for your responses everyone but I think I am WAY out of my depth here. Started off thinking I could choose a driver and chuck it in a cabinet made according to the volume specified by the manufacture and have a cracking sub (I knew that just sounded too easy). I have no understanding of the t/s parameters so not much of this makes sense to me. Does someone know some youtube tutorials or equivalent sources of education for noobs? I would love to know all of this stuff but I'm not sure how to learn it.
So could I construct a 1-1.3 cubic ft enclosure without doing a bunch of simulations I don't understand and have a great sounding sub? or would you recommend putting the time and effort into educating myself so I can do simulations? If so, a good start would be to know what I am aiming for within the simulations.
Thanks again!
Myles
So could I construct a 1-1.3 cubic ft enclosure without doing a bunch of simulations I don't understand and have a great sounding sub? or would you recommend putting the time and effort into educating myself so I can do simulations? If so, a good start would be to know what I am aiming for within the simulations.
Thanks again!
Myles
It is very surprising that you are negating tried, tested and prooved science of electrical-acoustical analogy by Thiele and Small. Critical damping is very real acoustical event, acoustically measured and subjectively heard in every possible case, without single one exception (that is the definition of a science fact).
The box do solves the acoustical back-front cancellation - how that can not be helpful?
Infinitely big box is wrong with every woofer in existence having TS parameters appropriate for a sealed box with acceptable dimensions. Everybody is disagreeing with you.
With 1-1.5 cubic ft sealed box you will have a great sounding sub.
Do not forget to put stuffing inside.
Last edited:
I would probably put it in a ~1 ft^3 (28 liter) box and forget about it. F3 will be perhaps 40Hz, which is reasonably deep for a sealed woofer.
A lot depends on intended use. If you plan to pound on this with close to rated power all the time, the 0.5 ft^3 box might be more appropriate.
If you want really low bass you can make a roughly 2 ft^3 vented box tuned to 23-24Hz or so (might need a PR) and get flat response down to the low 20's.
The electrical-acoustical analogy approach started long before Thiele* and is continuing to evolve today (for example, all the buzz about lossy inductance). Did you know that? Even people like me who think the Rice-Kellogg driver is a real stupid way to make sound, have no problem about that kind of model.
The question is how to build a speaker that reproduces sound faithfully and appropriate to the task in a room, not the validity of the model (which isn't materially contested by anybody).
Do you know anything about motional feedback? Any idea about the extremely low Q you can achieve that way? One reason it sounds so great.
Ben
*Thiele and Small were not scientific innovators but clarified the engineering approach some. Don't know why they are so deified on this forum. Pity Olson and Beranek are not given more credit.
Last edited:
Wrong ! It does more than just that, plus it's not a riddle, well maybe to you 😀 Most people with any sense, would find raising effeciency more than a little helpful 🙂 Horns/TH's/Reflex/BP etc boxes can all do that.
Yes, me & a lot of others too. The box has to be the Correct size etc, to fulfill a target goal. That's why Sims are important, as are the correct T/S specs, easily found by testing.
Of course, unless we all build an infinitely big box with motional feedback, we are all missing out 😛 And anyway, how could You go about Correctly designing a MF circuit/system without using component data & sims etc ? You couldn't ! Same goes for designing boxes of Any type.
Thiele and Small both credited others such as Olson & Beranek, as many of us realise.
Leach and Wright both had models to quantify lossy inductance effects and a way to extract lossy inductance t/s parameters DECADES ago. And it wasn't even really a problem back then. It's only an issue now with new class of super woofers with huge excursion potential.
This recent buzz that I started about lossy inductance is because:
1. Most people still don't know, they still think the sim is valid for this type of driver.
2. Lossy inductance t/s parameters are rarely available, you have to measure them yourself.
3. Unibox is the only simulator I know of that can accept these lossy inductance parameters.
4. I don't even know if the lossy inductance parameters will lead to a more accurate sim.
5. In light of all the above I needed a way to make a more accurate sim, so I did.
Yes, we all know you don't like moving coil drivers. And yet you have owned and continue to own ONLY moving coil drivers for subwoofers.
If they are so stupid why do you continue to use them?
If they are so stupid what is the alternative? You claim to have engineering experience yet you've never proposed an alternative solution. Instead you talk about (but have never used) motional feedback as the solution, which does nothing to replace the primary ingredient - the moving coil driver.
How can you possibly know how the sub will behave in room if you don't even know how it will behave outside? Most of us have a pretty good idea what's going to happen when you stick a speaker in a room.
Your solution has always been to just stick it in a large sealed box and that's as good as it can possibly get, but that is so wrong.
You still think that any driver in a large sealed box is going to have a "whomp up" at fs. Even though I've told you dozens of times this is not so unless the driver has VERY high qts, even though I've shown you sims to show why this is wrong, you still continue to spout this misconception. As recently as yesterday or the day before you said this "whomp up" would be seen in OB too, which is even more wrong UNLESS the driver has very high qts.
The simulator is VERY important, it shows a reality that lets us avoid silly mistakes like the ones you continue to post year after year.
For perhaps the 20th time now I'm telling you that the lowest possible q is not a good goal. You have even said that the ideal q is zero. You don't seem to understand what q is at all. With very low q you have very little bass, if the q is low enough you have no bass at all. You certainly won't get a "whomp up" at fs with a very low q.
You have demonstrated a lack of understanding of the very basics of audio theory and how stuff works and have continued to post these misconceptions for years now. You refuse to even look at anything that disagrees with your preconceived (wrong) ideas about moving coil drivers, resonances, q, etc. It is extremely tedious constantly correcting your mistakes but until you learn how this stuff actually works we will have to continue to correct you.
Stuff that hasn't been mentioned here -
Box qtc describes frequency response, how the low knee will look.
Low qtc = shallow rolloff starting at a very high frequency
Qtc = .707 = max flat = flattest possible response to the lowest possible frequency
High qtc = bump in response at resonance and then a steep rolloff
As far as I'm concerned qtc doesn't matter much. Qtc describes frequency response and frequency response can (and should) be eq'ed. Furthermore, to achieve any substantial reduction in qtc you need to make the box quite a bit bigger and you will still have to eq.
The FAR more pressing matter is power handling and how much power is required to reach excursion limits. See those tests on data-bass? He's plowing those drivers with something like 5000 watts. This is necessary because he uses very small test boxes so it takes massive power to reach excursion limits. It should be fairly obvious that 5000 watts into a driver is probably not the best way to run the system. There's very few amps that can do that much power (the improved 14000 watt Chinese clone amp can only do 3800 watts sustained and my Behringer EP4000 can only do about 1400 watts sustained) and even if you could give your sub 5000 watts there would be massive compression and huge risk of thermal failure.
Making the sealed box larger gives you better efficiency at lower frequencies where you need it and can reduce the amount of power required to reach excursion limits by 10x. So you only need a 500 watt amp instead of 5000 and there's no risk of thermal failure at all and unlikely power compression will be an issue either.
Any decent simulator can show all of this very clearly.
But sometimes small size is a primary goal and there's nothing wrong with that. Some people want a compact sub. They should realize the pros and cons of the situation and choose what is best for them.
Box qtc describes frequency response, how the low knee will look.
Low qtc = shallow rolloff starting at a very high frequency
Qtc = .707 = max flat = flattest possible response to the lowest possible frequency
High qtc = bump in response at resonance and then a steep rolloff
As far as I'm concerned qtc doesn't matter much. Qtc describes frequency response and frequency response can (and should) be eq'ed. Furthermore, to achieve any substantial reduction in qtc you need to make the box quite a bit bigger and you will still have to eq.
The FAR more pressing matter is power handling and how much power is required to reach excursion limits. See those tests on data-bass? He's plowing those drivers with something like 5000 watts. This is necessary because he uses very small test boxes so it takes massive power to reach excursion limits. It should be fairly obvious that 5000 watts into a driver is probably not the best way to run the system. There's very few amps that can do that much power (the improved 14000 watt Chinese clone amp can only do 3800 watts sustained and my Behringer EP4000 can only do about 1400 watts sustained) and even if you could give your sub 5000 watts there would be massive compression and huge risk of thermal failure.
Making the sealed box larger gives you better efficiency at lower frequencies where you need it and can reduce the amount of power required to reach excursion limits by 10x. So you only need a 500 watt amp instead of 5000 and there's no risk of thermal failure at all and unlikely power compression will be an issue either.
Any decent simulator can show all of this very clearly.
But sometimes small size is a primary goal and there's nothing wrong with that. Some people want a compact sub. They should realize the pros and cons of the situation and choose what is best for them.
Well, if you only need 50 watts of power, just think of how much cooler the voice coil will be, and the extra energy not lost to IR losses, never mind the parameter shift! In short, to create the best low frequency bass, the box should be the correct size for the woofer, eliminating the need for EQ and giant amplifiers. A B4 ported box sounds fantastic as ling as the fs of the woofer is low enough and your Qts is around 0.383. Sealed boxes, transmission lines and horn enclosures each make the most of various woofers. As a very rough guide, a B4 ported will be about 2X the size of a sealed.
The driver really dictates the type of box you build if you want to get the most out of it. Work with physics rather than against it for the most efficient and best sounding speaker system. The one technology that attempts to defy the rules of physics are motional feedback systems in tiny boxes (like most subwoofers today). They also rank #1 in failures. That shouldn't be a surprise to anyone. For ported systems, keep in mind that below the system resonant frequency, the port unloads the woofer and then you will easily reach past the limits of excursion and damage. That's why you want woofers that have a really low resonant frequency with those.
-Chris
The driver really dictates the type of box you build if you want to get the most out of it. Work with physics rather than against it for the most efficient and best sounding speaker system. The one technology that attempts to defy the rules of physics are motional feedback systems in tiny boxes (like most subwoofers today). They also rank #1 in failures. That shouldn't be a surprise to anyone. For ported systems, keep in mind that below the system resonant frequency, the port unloads the woofer and then you will easily reach past the limits of excursion and damage. That's why you want woofers that have a really low resonant frequency with those.
-Chris
Back to topic please! It is possible to respectfully disagree
So I have a fairly large range of sizes that will work well with this driver, good 🙂 Ported is out of the picture at the moment. Don't even want to know about it until I've successfully built a sealed sub. The sub will be used to accompany my bookshelf speakers in a in a fairly small bedroom for the moment. I will be moving house soon and have no idea what room they'll be in next so versatility is key.
I am quite possibly wrong and obviously there is some debate around this but is it safer to err on the larger side as it may give a better bass response? So a 1.3 ft^3 enclosure with some stuffing should go down a treat? Do I need to account for the volume taken up by the driver? and do I need to brace the walls or is it small enough that I don't need it?
Cheers 🙂
Myles
So I have a fairly large range of sizes that will work well with this driver, good 🙂 Ported is out of the picture at the moment. Don't even want to know about it until I've successfully built a sealed sub. The sub will be used to accompany my bookshelf speakers in a in a fairly small bedroom for the moment. I will be moving house soon and have no idea what room they'll be in next so versatility is key.
I am quite possibly wrong and obviously there is some debate around this but is it safer to err on the larger side as it may give a better bass response? So a 1.3 ft^3 enclosure with some stuffing should go down a treat? Do I need to account for the volume taken up by the driver? and do I need to brace the walls or is it small enough that I don't need it?
Cheers 🙂
Myles
Last edited:
What I would do is set some goals and then look at some sims of various sizes of sealed box. Check the pros and cons, how much power does the driver need to reach it's limits and can the driver handle that much? What's the frequency response like?
When you look at a range of sizes you will see that there are HUGE changes when the box is tiny vs when it's much larger than Vas. And you will notice that small changes in volume don't have much effect.
Is it worth it to go from 1.3 to 2 ft? Is it worth it to go all the way up to IB? Only you can answer that, hopefully with a well defined list of goals and some insight into how they will perform (sims).
Yes you need to account for the space the driver takes up. Yes you need bracing.
Don't forget the volume taken by bracing and anything else in there (such as a port in vented systems). You might be surprised by how much room bracing can take up.
-Chris
-Chris
Just looking at sims now and there really isn't much diff between say 1.2 and 1.6 ft^3 in terms of transfer function you're right.
When I do sims for cone excursion with system input power at 350W (the drivers RMS) it easily exceeds the Xmax in a 1.5 ft^3 box. Isn't this something I should be concerned about? Ill probably very rarely supply that much power to it but even at 150W in a 1.5 ft^3 box the cone excursion reaches xmax on the lower frequencies. Ill attach some screenshots.
1.5 ft^3 at 350W
1.5 ft^3 at 150W
0.6 ft^3 at 350W
What other sims should I do and what should I look for in the graphs?
Thanks!!
Let me suggest that it would help you to understand a bit better how these various simulation models work. There are quite a few different simulation programs, but most of them are based on fairly simple electrical models that have been well documented by Thiele and Small and J. E. Benson. Recently Jeff Candy provided a nice tutorial on the driver/cabinet models. It's math-intensive because Jeff has a PhD in physics or some related field, but if you just try to follow the concepts I think you will get an idea of what's "under the hood" for these models. Once you understand that, these different simulation programs and their results might not appear as confusing.
Jeff's excellent tutorial is here: http://techtalk.parts-express.com/filedata/fetch?id=1274483
- Status
- Not open for further replies.