Sub enclosure design - calling any experts...

B

bryston

Guest
I fancy the idea of a little DIY project constructing a sub. Although I am handy at construction I have no idea where to start with regard to the design. If there is anyone on here who can share some expertise I'd appreciate it! I don't really even need a sub although I think it could be fun making one and learning a thing or two along the way.

The sub will be used in a spare room (only a 14x10') rather than the lounge initially. Application is for two channel music, with the odd film via a plasma. The only priority is sound quality. It must be as accurate as possible, so as not to drag down my hifi into blurred sub bass. Would be good to limit max size to 8 cubic foot. (Kit is benchmark dac1 into pmc aml1 active near field monitors).

Considerations:

1. Should I use a large driver that only has to move a tiny amount (say a Precision Devices 24") or a smaller driver (JBL pro?) that would need to move a lot more to shift the same amount of air. Discussion most welcome! Remember accuracy, not spl is important.

2. Transmission line designs are the most accurate I've ever experienced so far. I'm open to other solutions though - sealed, ported, infinite baffle (although not really practicle in this room, but still interesting). Any thoughts?

3. Power could be the usual suspects of Bryston or Crown mono's. I'd like to hear some other ideas though. Perhaps digital or DIY options?

4. Cross over could be a Behringer, or...? I was thinking leave the floor standers running a full signal (which they do very well) with the sub only joining in from 15-40hz ish? Methods appreciated!

5. Location - would be good to have it placed bang in the centre of the room, as a single sub solution. Acting as a coffee table too :) I understand 2 or 4 subs in the corners (or moved out a little) is another ideal solution. However I'd like to just make one sub.

I was looking for free downloadable sub enclosure design software. So far I have these links below. Any reccommendations? A link to designing would be a good start for me too! Thanks in anticipation...

http://www.ajdesigner.com/speaker/index.php

http://www.caraudiohelp.com/subwoofer_enclosure_software/subwoofer_enclosure_software.htm
 
It certainly looks that way, but we'd be interested to see and hear about your results. Best of luck.

Russell
 
I missed this post earlier, or would have replied sooner ...
This is going to be long, and has some obscure parts where I am deliberately too brief. Some diligent searches should help - everything you need to know is freely available on the web (I checked a few months ago).

"If there is anyone on here who can share some expertise I'd appreciate it!"
OK I'll have a go - having made a few subs for myself and others. I should declare, however, that last time, after much effort designing, I found I had something very very Monolith like, and since there was no chance of me building it so well, I just bought one... (or soon two)

"The sub will be used in a spare room (only a 14x10') rather than the lounge initially."
The bigger you can go with the sub, the easier the design will be, if you aim for a given LF cutoff without too much distortion that is. You may be told that a big sub is bad for a small room - and that may be true if you don't use any EQ, but with EQ it is better to cut away excess than boost what is not there in the first place. And there is nothing wrong with EQ!

My attitude to sub design is to minimise distortion, and sort out the response with EQ (you'll be able to design a sub with a better anechoic response than the actual in room response - I'm not advising starting with a poor response).

"Application is for two channel music, with the odd film via a plasma. The only priority is sound quality.

If you have decent EQ you can switch from an exaggerated "movie" mode to a more normal balance for music. (Note that will not usually be a flat response, but a "house curve" gently rising to LF: search here and elsewhere)

"It must be as accurate as possible, so as not to drag down my hifi into blurred sub bass. Would be good to limit max size to 8 cubic foot. (Kit is benchmark dac1 into pmc aml1 active near field monitors)."

The design factors are

distortion as a function of level - a high efficiency acoustic design combined
with a driver with a low distortion motor is the starting point (only a true domestic sub driver is going to make a good domestic sub - there are very few available in the UK unless you self import, as usual more in the US. At 8 cubic feet forget the Peerless ones).

transient response or frequency/phase response of the EQ'd sub in-room (two ways of expressing the same information) you can forget all the "nonsense words" like "fast bass", "muddy bass" etc. Measure and be done with it - no sales talk needed with DIY.

"Considerations:

1. Should I use a large driver that only has to move a tiny amount (say a Precision Devices 24") or a smaller driver (JBL pro?) that would need to move a lot more to shift the same amount of air. Discussion most welcome! Remember accuracy, not spl is important."

Accuracy of what? I don't quite follow your question. EQ patience can sort out the frequency response, and sound design will keep distortion low, but that is a strong function of SPL. Big drives have large inductances in the motor (for given SPL/W) and that hurts transient response - several small ones in parallel wins there.

The cone also has to hold together up to somewhat above your selected crossover - that gets harder for big drives, explaining most of the cost.

Because of these two considerations and driver availability I favour multiple 12" drives over anything very much larger, but that is mainly a value for money argument and not anything absolute. I have no direct experience with larger units, nor would I want to have (except if making an IB where a few 15" designs are fine, as I'm sure Nimby can confirm).


"2. Transmission line designs are the most accurate I've ever experienced so far. I'm open to other solutions though - sealed, ported, infinite baffle (although not really practicle in this room, but still interesting). Any thoughts?"
My most satisfactory "sub" design was a TL, but not one that went very low (it was extending very small mains down to ~30 Hz). TLs are probably the least efficient approach in terms of distortion compared to volume (cubic feet), and get very hard to make work with a very low cutoff. The 30 Hz ones had about the same volume as a Monolith, and more distortion too!


3. Power could be the usual suspects of Bryston or Crown mono's. I'd like to hear some other ideas though. Perhaps digital or DIY options?
BKelec modules? Hard to beat IMHO. I've built amps up to 300W, and it is really not worth it with BKelec prices. Alternatives are Hypex modules, if you want efficiency. Except for truely pathological cases there is no such thing as sound quality for sub amps. If SPL is not that important a few hundred W should suffice in the room you mention.

"4. Cross over could be a Behringer, or...? I was thinking leave the floor standers running a full signal (which they do very well) with the sub only joining in from 15-40hz ish? Methods appreciated!"
Ah, here I see the whole purpose of a sub is to get bass without too much distortion, so high-passing the mains is a must! (Just an opinion; bass distortion is so obvious.) I do it using DSP in a PC with an excellent sound interface to obtain perfect 6th order filters at 100 Hz (my electrostatic speakers have terrible distortion with any real bass though). A Behringer crossover can do it too (the fancy DSP one that has been hard to get recently) - and can be used with the BFD or other parametric EQ to get the response you want.

"5. Location - would be good to have it placed bang in the centre of the room, as a single sub solution. Acting as a coffee table too :) I understand 2 or 4 subs in the corners (or moved out a little) is another ideal solution. However I'd like to just make one sub."
Actually for many rooms the middle of a wall gives the best result for a wide range of listeners, the corners for one listener (I can't find the reference but there is a great web page showing examples). With one sub, I put it behind the mains at the back of the room, use a time delay correction to ensure phase coherence. The coffee table is going to be very locatable (localisable?) unless you really work on the steep crossover and low distortion.

"I was looking for free downloadable sub enclosure design software. So far I have these links below. Any reccommendations? A link to designing would be a good start for me too! Thanks in anticipation..."

I suggest sticking to basic bass reflex - why? Sealed box is "simpler", but less efficient for a given volume of box and target frequency, also there are not too many drives that work great in a big sealed box (that I found anyway). Bass reflex gives a good balance of performance against volume, and if EQ is assumed slight errors of design are unimportant as they can be corrected (remember only the net frequency response matters, all of this stuff in the box is minimum phase so the parametric EQ can compensate speaker resonances accurately). The design equations are not too complicated (it is just a mass on a spring, with damping for the drive, coupled to the box resonance). I usually just bung the equations into MATLAB or similar, but there are many alternatives that give the same answer ("boxplot" seems to stick in my mind as one that a friend used).
As I'm sure you know you need the drive parameters that describe the resonance and Q. You then select a box volume and port size/tuning to get a flattish well extended response. To gain experience I suggest starting from a known design, either commercial or a project published on the web. You could start with the Emminence drive used (in a different version) in the Monolith, in 100 l box, with about the same size of port (you can work it out from the pictures and box dimensions - the port goes up to a couple of inches from the back). Get a feel for what parameters affect what, and soon you'll be able to look around for drives that offer something you want.
Design software that allows you to tell if the ports are too small is worthwhile (as port noises are horrible) - on the other hand look at comparable designs.
A poor drive will force you to use a port longer than a sensibly shaped box.

The box should be perfectly sealed, sensibly shaped (look at decent subs) and, of course, rigid.

If you get to the stage where the design looks OK, but you want more than just a 12" - you can always double the drives, volume etc. But also consider multiple smaller subs (I'm soon going to add my second Monolith when I move to a larger place in some weeks). Of course setting up multiple subs is a challenge - but that is what you want, right? (EQ is enough of a challenge for me these days.)

Have considerable fun!
Ken
 
It has been said that no matter how clever you are at DIY, or however experienced at designing subs, you just cannot make a better sub than SVS for the price they charge.

I think Ron said that. :D
 
Definintely so!

Even the parts costs are comparable (because of huge margins taken by most parts suppliers), and then there is the time - learning, design and sourcing parts is likely to take over 100 hours.

Ken

Nimby said:
It has been said that no matter how clever you are at DIY, or however experienced at designing subs, you just cannot make a better sub than SVS for the price they charge.

I think Ron said that. :D
 
Kstrain, thanks for the detailed reply. Certainly a lot of info there.

I was looking at drivers with a low Fs and high Qts, so I can achieve real subsonics. The volt 15" looks very promising. However I also found an 18" as used in car audio which will give an F3/-3db of 22Hz with a small box. That sounds rather deep :)

I did wonder if it might be better to just buy a velodyn last night. :rolleyes:
 
What you get with a bought sub is a well tested example of the best the designer(s) can mange using extraordinary investment in time and materials.

Whether you like what they come up with (and then often have manufactured in a low-wage ecomomy to maximise profit thoughout the distribution chain) is a matter of opinion. One not always shared by all.

With respect, your knowledge of subwoofers does not make you an ideal candidate for success at manufacturing your own subwoofer. The chances of success are probably so slim as to be a complete waste of your time and money. You may learn more than you do now but your almost certainly will not have a subwoofer that matches even a modest AV setup let alone one of your own standards.

If you absolutely must build a sub do it as an experiment rather than seeking perfection. I have been building AV stuff (amongst many other things) for more than 4 decades. If I felt driven to make a decent subwoofer I would copy a succesful example of a DIY sub. One where the response curve is published and discussed on a technical forum. I would make the copy as close as humanly possible with the skills, tools and materials I had available. Then I would hope that I had succeeded in my copying given the variablity of drivers and cabinet manufacturing skills and conditions.

Go out a buy a decent sub. You can afford it. Make your dream sub between film or music sessions and be prepared to laugh at yourself for even trying.
 
bryston said:
Kstrain, thanks for the detailed reply. Certainly a lot of info there.

I was looking at drivers with a low Fs and high Qts, so I can achieve real subsonics. The volt 15" looks very promising. However I also found an 18" as used in car audio which will give an F3/-3db of 22Hz with a small box. That sounds rather deep :)

I did wonder if it might be better to just buy a velodyn last night. :rolleyes:
I hope I say this without being patronising, remember I don't have a clue how much you know (you could be a physics professor like me)! But both of the statements you make above suggest you have a huge amount to learn about subwoofer design.

The standard drive parameters "disguise" the physics behind them (apart from mass), so you cannot tell that a drive is going to be good from Fs and Qts - e.g. the ratio of the electrical and mechanical Qs tells much. Probably you need ~2nd year univeristy physics (or equiv.) to make this intuitive. Without that it will be trial and error (fun or frustrating, and probably expensive). I think your choice of starting point is not a good one.

In car responses tell almost nothing - unless your room is car sized, in a perfectly sealed car the response can go down to 0 Hz in principle.

I think using the Volt unit would be a mistake, if I interpret what your needs are. Those drives are meant for another purpose. Take a look at units in commercial subs.

If you are determined everything is free on he net, and better fora exist where you'll have a chance of getting help from people who still build subs.

I have to agree with Nimby, it makes much sense to buy a suitable number of Monoliths or (larger boxed) SVS subs and the tools for EQing and you can get nearly as much fun and frustration.

Still, good luck,
Ken
 
I enjoy making things and I thought it might be achievable given enough time. Perhaps I underestimated quite how much time so I'll take your advice afterall - buy a solution and just read up on the theory in my spare time. Thanks guys. :lesson:
 
You're not convinced by the IB route to subwoofer heaven?

No real skill required except cutting the driver holes with an electric router.

You should get decent SQ and extension even without equalisation depending on your choice of driver. You'll probably have to import 4 x 15" (or 18") drivers from the USA or Canada. I was given a very reasonable quote from a Canadian dealer who sells a useful 18" driver on Ebay for absolute peanuts. The freight charges were too steep for my tastes and I didn't really need bigger drivers anyway. I'd like four more 15" inchers like the first four. (Temporarily not being manufactured)

Use a manifold and you can hide it under the floor if you have enough space. Or in the roof if your listening room is under the roof. In the wall if you have an attached garage or workshop. Spare bedroom adjoining wall?

See gallery and illustrated FAQ for inspiration:

http://ibsubwoofers.proboards51.com/index.cgi?
 
kstrain said:
it makes much sense to buy a suitable number of Monoliths or (larger boxed) SVS subs

Boxed? Is this because the cylinders are not available in UK?
I like the idea of the cylinders, for style and lack of kg.
 
nwhitta said:
Boxed? Is this because the cylinders are not available in UK?
I like the idea of the cylinders, for style and lack of kg.

sorry I meant "large-boxed", or better, "large-volume". Any shape will do
(but it seems that cylinders are hard to get in the UK).

Ken
 
Nimby said:
You're not convinced by the IB route to subwoofer heaven?
Now you've got me thinking - I wonder what the attic space is like in my new house (not moved yet). Dare I start cutting holes in the ceiling ... then again the utility room is next to the lounge at one end, and has nothing particularly useful in it, dare I cut through the wall ....

The $ rate is good for buying from the US too, oh my.

Perhaps another Monolith is less likely to lead to regret ....

Ken
 
Ken

You don't have to make a large hole in the wall.

Use the garage for an apposed-driver, box manifold. Takes up hardly any room.

No more boxes indoors and bass where the word "distortion" has no meaning.
 
Even a small hole concerns me a little (esp in the ceiling, even with a grill), but the idea is very tempting. I could get someone to help with the hole of course, and I can manage the wood for the manifold OK. No suitable garage, unfortunately, but there are other possibilities. Hmm....

Thanks for the encouragement.

Ken


Nimby said:
Ken

You don't have to make a large hole in the wall.

Use the garage for an apposed-driver, box manifold. Takes up hardly any room.

No more boxes indoors and bass where the word "distortion" has no meaning.
 
I feel your pain, Ken.

I wasn't "allowed" to make holes in the ceiling either. :blush:

The edge of the floor would be favourite if you have a suspended wooden floor with some breathing space below.

At least a foot or two of crawl space is desirable. If there isn't much clearance where you want the manifold then you can excavate a little and sneak out the excess soil in a vaulting horse. ;)
 

The latest video from AVForums

TV Buying Guide - Which TV Is Best For You?
Subscribe to our YouTube channel
Back
Top Bottom