Hi,
I am looking at building a subwoofer, but am a little unsure of a few things. Hopefully someone here can help.
If I had 2 subs to choose from. One being larger in diameter, but smaller xmax, both with the same piston volume. Which would be ther better choice? why? All other things considered equal.
I am after low bass more so than brute SPL. I am considering using the Adire Tempest in the 214L vented enclosure. But from the few models I have messed with on LspCAD I can get smaller larger throw drivers to extend lower. Is it purely SPL where the larger driver will prevail?
Last question is about frequency response. Any tips to help avoid boominess at low freqs. I read that you should avoid a curve that is flat all the way to 20hz. Due to room gain.
Any input would be great.
Thanks
Regards
Jamie Wilson
I am looking at building a subwoofer, but am a little unsure of a few things. Hopefully someone here can help.
If I had 2 subs to choose from. One being larger in diameter, but smaller xmax, both with the same piston volume. Which would be ther better choice? why? All other things considered equal.
I am after low bass more so than brute SPL. I am considering using the Adire Tempest in the 214L vented enclosure. But from the few models I have messed with on LspCAD I can get smaller larger throw drivers to extend lower. Is it purely SPL where the larger driver will prevail?
Last question is about frequency response. Any tips to help avoid boominess at low freqs. I read that you should avoid a curve that is flat all the way to 20hz. Due to room gain.
Any input would be great.
Thanks
Regards
Jamie Wilson
Its not the matter of big dia. or long exc. but the volume of air your driver can push around.
If you want something 20-30 hz at 0db your "generally" have to go with bigger driver(bigger Vas - enclosure will be bigger, thus it'll be easier to tune it lower, futhermore smaller eclosures tend to "lag"(group delay drastically increases as frequency decreases) if tuned low)
If you want something 20-30 hz at 0db your "generally" have to go with bigger driver(bigger Vas - enclosure will be bigger, thus it'll be easier to tune it lower, futhermore smaller eclosures tend to "lag"(group delay drastically increases as frequency decreases) if tuned low)
VEC7OR's comments about air volume are right on the mark.
Other thoughts:
Consider how high you want to take the sub. Higher lowpass frequencies may call for smaller diameter drivers.
Long throws are nice, but nonlinearities are often a problem. Large dameters are nice, but beware cone breakup if you try extending the frequency range upward.
Multiple smaller diameter drivers can often be a good compromise.
Room gain is real and often (usually) unpredictable. EQ is not to be feared- it can be your best friend.
Other thoughts:
Consider how high you want to take the sub. Higher lowpass frequencies may call for smaller diameter drivers.
Long throws are nice, but nonlinearities are often a problem. Large dameters are nice, but beware cone breakup if you try extending the frequency range upward.
Multiple smaller diameter drivers can often be a good compromise.
Room gain is real and often (usually) unpredictable. EQ is not to be feared- it can be your best friend.
Thanks guys for your insight.
To clarify a couple things.
1. Looking for a good response from 80-20 anything below that is a bonus.
2. Need a box no bigger than 200L
3. Will be using HS500 amp. 700wrms into 4ohm.
What are nonlinearities? What do they SOUND like??
OK, and group delays are bad why?
Boy what a learning curve!
Any particular drivers I should consider given my requirement?
Thanks again.
To clarify a couple things.
1. Looking for a good response from 80-20 anything below that is a bonus.
2. Need a box no bigger than 200L
3. Will be using HS500 amp. 700wrms into 4ohm.
What are nonlinearities? What do they SOUND like??
OK, and group delays are bad why?
Boy what a learning curve!
Any particular drivers I should consider given my requirement?
Thanks again.
Hi Wilson,
I have a Tempest in a 250l vented box, it sounds very good, but a box this big you will need to brace so dam well for it to stay even nearly still.
I have a 150w/4ohm plate amp and it goes very loud. The HS500 is probably a little bit overkill, but certainly a fine choice, I bet it will sound great with all that headroom, and probably damping factor - to keep things tight.
One point though, where do your speakers roll off at, at -3db? My kit speakers supposedly do so at around 45-50hz, and 50hz is my minumum cut-off for the sub. This creates slightly too much overlap and makes things just a tad thick in the region. It is worth considering your x-over requirements, before buying an expensive plate amp.
I think group delay is the phase difference/time lag between parts of the audio band reproduced. Presumably the lower in frequency the problem area is the better - ie. one reason a nice low tuning frequency is better than a higher one (where ears are more sensitive). Tuning with a port can cause that excessively detached bloated bass around the tuning frequency and I think group delay has something to do with it.
One more point - will two Shivas (ie. 12" version of Tempest) not be far better than one Tempest? Use them in a push-push arrangement, so firing apart from each other. Iirc, 2 Shivas only need roughly the same space as 1 Tempest, and I bet they sound tighter, because they will have more magnetic force per cone mass. Furthermore, operating in a push-push stylee will cancel some vibrations - useful in a huuge box like we have.
-Simon
I have a Tempest in a 250l vented box, it sounds very good, but a box this big you will need to brace so dam well for it to stay even nearly still.
I have a 150w/4ohm plate amp and it goes very loud. The HS500 is probably a little bit overkill, but certainly a fine choice, I bet it will sound great with all that headroom, and probably damping factor - to keep things tight.
One point though, where do your speakers roll off at, at -3db? My kit speakers supposedly do so at around 45-50hz, and 50hz is my minumum cut-off for the sub. This creates slightly too much overlap and makes things just a tad thick in the region. It is worth considering your x-over requirements, before buying an expensive plate amp.
I think group delay is the phase difference/time lag between parts of the audio band reproduced. Presumably the lower in frequency the problem area is the better - ie. one reason a nice low tuning frequency is better than a higher one (where ears are more sensitive). Tuning with a port can cause that excessively detached bloated bass around the tuning frequency and I think group delay has something to do with it.
One more point - will two Shivas (ie. 12" version of Tempest) not be far better than one Tempest? Use them in a push-push arrangement, so firing apart from each other. Iirc, 2 Shivas only need roughly the same space as 1 Tempest, and I bet they sound tighter, because they will have more magnetic force per cone mass. Furthermore, operating in a push-push stylee will cancel some vibrations - useful in a huuge box like we have.
-Simon
Group delay
Group delay is the time by which frequencies within a group are delayed. If you had a group delay of 1s for all audio frequencies, it wouldn't matter (except for when cueing the stylus). There's only a problem when you have non-constant group delay - some frequencies are delayed by more than others. Filters are not only characterised by their amplitude vs frequency response, but also by their delay vs frequency response. The low-pass filter needed for a sub-woofer always implies substantial delay, and unless you also delay the audio to the main loudspeakers, the sub-woofer will always lag behind. Additionally, the LF alignment of the sub-woofer and its enclosure forms a high-pass filter and as as the volume of the box decreases relative to VAS, Q rises and delay between different frequencies increases. There's nothing like a big box.
A larger diameter driver can move the same volume of air as a small diameter driver, but with less excursion. Less required excursion means the coil doesn't need to be so overhung, and that increases efficiency. A larger cone will break up at a lower frequency because break-up occurs when the circumference of the cone is equal to one wavelength. The velocity of sound varies with cone material, but break-up frequency is broadly inversely proportional to cone diameter. You have to try quite hard to get a cone to break up at <200Hz.
Group delay is the time by which frequencies within a group are delayed. If you had a group delay of 1s for all audio frequencies, it wouldn't matter (except for when cueing the stylus). There's only a problem when you have non-constant group delay - some frequencies are delayed by more than others. Filters are not only characterised by their amplitude vs frequency response, but also by their delay vs frequency response. The low-pass filter needed for a sub-woofer always implies substantial delay, and unless you also delay the audio to the main loudspeakers, the sub-woofer will always lag behind. Additionally, the LF alignment of the sub-woofer and its enclosure forms a high-pass filter and as as the volume of the box decreases relative to VAS, Q rises and delay between different frequencies increases. There's nothing like a big box.
A larger diameter driver can move the same volume of air as a small diameter driver, but with less excursion. Less required excursion means the coil doesn't need to be so overhung, and that increases efficiency. A larger cone will break up at a lower frequency because break-up occurs when the circumference of the cone is equal to one wavelength. The velocity of sound varies with cone material, but break-up frequency is broadly inversely proportional to cone diameter. You have to try quite hard to get a cone to break up at <200Hz.
If box size isn't an issue, I would pick the larger diameter speaker. As has already been mentioned, the larger driver will probably be more linear. What hasn't been mentioned is that the larger driver will also be more efficient than the smaller counterpart. You will be able to use a smaller amp to push it to xmax. These are however generalizations. There always seem to be exceptions...
This is my uneducated understandig on nonlinearities and compression, so everyone feel free to correct me.
If you play a test tone at 1 watt, your speaker will play that tone at some volume, say 80db. Double the power to 2 watts, and the speaker should play at 83db. Double again to 4 watts, and it will play at 86db.
Now imagine you are playing the tone at 200 watts and getting 100db. If you double the power to 400 watts, the speaker should play at 103db. But because you are nearing the capabilities of the driver, it may play the tone at 102.8db. At higher power levels 'compression' becomes a increasingly important factor. Doubling in power does not result in 3db gain. The way you will hear this is by noticing a reduction in the dynamics of the music.
Hope that helped...
Dan
This is my uneducated understandig on nonlinearities and compression, so everyone feel free to correct me.
If you play a test tone at 1 watt, your speaker will play that tone at some volume, say 80db. Double the power to 2 watts, and the speaker should play at 83db. Double again to 4 watts, and it will play at 86db.
Now imagine you are playing the tone at 200 watts and getting 100db. If you double the power to 400 watts, the speaker should play at 103db. But because you are nearing the capabilities of the driver, it may play the tone at 102.8db. At higher power levels 'compression' becomes a increasingly important factor. Doubling in power does not result in 3db gain. The way you will hear this is by noticing a reduction in the dynamics of the music.
Hope that helped...
Dan
I vote for diameter every time plus very big magnet assembly. If you look at driver linearity it is only reasonably so over an excursion of a few milimeters. Undersized subwoofers are not nice and should only be used if there is no other option.
Remember cone area increases by r squared so even relatively small increases in cone diameter result in substantial increases in area.
Remember cone area increases by r squared so even relatively small increases in cone diameter result in substantial increases in area.
Think Volume and not diameter or excursion
You need alot of air to even hear the lower frequencies. You need approximately 2 litres of real not hopefull displacement to hear 20 hz. The lowest distorsion will come from a bunch of smaller diameter woofers. To figure out real x-max you need to figure out voice coil length minus air gap height then add 25%. That will give you real X-max. You will find that the diameter of a woofer is usually usefull to about half the surround width as a true piston. With this info do some more looking and you may be surprised at the volumes that you come up with compared to the advertised specs. One more note. To hear 16hz at the loudest levels I have ever heard of you need four 15" woofers displacing 4litres of air per stroke. Not chicken feed and pretty darn impressive if you have ever heard it.
Mark
You need alot of air to even hear the lower frequencies. You need approximately 2 litres of real not hopefull displacement to hear 20 hz. The lowest distorsion will come from a bunch of smaller diameter woofers. To figure out real x-max you need to figure out voice coil length minus air gap height then add 25%. That will give you real X-max. You will find that the diameter of a woofer is usually usefull to about half the surround width as a true piston. With this info do some more looking and you may be surprised at the volumes that you come up with compared to the advertised specs. One more note. To hear 16hz at the loudest levels I have ever heard of you need four 15" woofers displacing 4litres of air per stroke. Not chicken feed and pretty darn impressive if you have ever heard it.
Mark
Four 15" sub drivers, presumably in a ported box. Yikes! That's one large cabinet! Did it make you feel ill?
It's a shame the laws of physics are the way they are
I'd love a little sub with a 10" driver or something like that, but it's just not gonna do squat. Ah well...
I still think 2x12" has to be a great compromise - for price, size, performance, versatility even(??)
The REL Studio II uses a couple of Volt 10 or 12 inchers. I bet that sounds nice. Has anyone heard volt bass drivers, or such a costly sub?
Has anyone tried their ARM (acoustic resistive matrix iirc) loading in a DIY project??
I've just remembered, some of the most impressive deep bass I've heard. Our city's cathedral is having a new organ built, costing millions £££. In the meantime they have a Hammond organ and large arrays of speakers on the walls. In one corner is a giant transmission line with 8(iirc) 15"(most likely, it's quite high up so could be 18") drivers!!! Sure it's a large building, but it can wobble the pews, and if you're feeling a little fragile after a big Saturday night, it makes you feel funny with the right hymns!!
-Simon
Following is a chart I posted for how much air must be moved to produce which SPL at what frequency. The original chart is by Richard Small. There are two charts-one for real and one for Metric measurements.
http://www.diyaudio.com/forums/showthread.php?threadid=5668&highlight=SPL+displacement+chart
Just for reference sake, the air a woofer can move is it's area times it's ±excursion. So a 15 woofer normally has something like 132 in² and an excursion of anywhere between ±¼" and ±1½", (yes, the latter actually exist). ±½" is a nice long excursion, so 132 in² X ½" equals 61 in³.
According to our chart that will give us about 104 dB @20 Hz. Four of them will raise the SPL 12 dB, so it will be 116 dB @ 20 Hz.
This applies to sealed systems only. Properly configured ported systems with the correct woofer parameters can give a single ported speaker the same output as four sealed woofers.
http://www.diyaudio.com/forums/showthread.php?threadid=5668&highlight=SPL+displacement+chart
Just for reference sake, the air a woofer can move is it's area times it's ±excursion. So a 15 woofer normally has something like 132 in² and an excursion of anywhere between ±¼" and ±1½", (yes, the latter actually exist). ±½" is a nice long excursion, so 132 in² X ½" equals 61 in³.
According to our chart that will give us about 104 dB @20 Hz. Four of them will raise the SPL 12 dB, so it will be 116 dB @ 20 Hz.
This applies to sealed systems only. Properly configured ported systems with the correct woofer parameters can give a single ported speaker the same output as four sealed woofers.
I think it sounds better when a 15" pushes air gently at say 30 Hz than to get the same SPL from a 12" trying get the same result at close to double the Xmax. I have realised this before I ever read it on DIY Audio sites.
I'm not sure what group delay is exactly or a lot of speaker terms. I think Ron E's quote somewhat applies here (no offense to any one ): "An expert is one who knows more and more about less and less until he knows absolutely everything about nothing. - Nicholas Murray Butler"
It's just simple physics, that moving air gently is a good idea. Depends what spl you want and at what Hz. Deep and loud bass on one 15" always sounds better to me than one 12" working twice as hard. Less SPL or more Hz a 10 or 12 might sound better to me
. Okay maybe the guy is a bit too much
. You know what I'm trying to say though
I'm not sure what group delay is exactly or a lot of speaker terms. I think Ron E's quote somewhat applies here (no offense to any one
): "An expert is one who knows more and more about less and less until he knows absolutely everything about nothing. - Nicholas Murray Butler"It's just simple physics, that moving air gently is a good idea. Depends what spl you want and at what Hz. Deep and loud bass on one 15" always sounds better to me than one 12" working twice as hard. Less SPL or more Hz a 10 or 12 might sound better to me
. Okay maybe the guy is a bit too much. You know what I'm trying to say thoughJimmy:
I kind of agree that bass from a larger woofer sounds better. But the larger the woofer, the bigger the box for the same cutoff. So it is a tradeoff.
I am glad that the poster came forward to explain group delay. It is a term that is getting increasing attention, an I am glad that someone came forward to tell us what it is. If you don't "get" the explanation, or simply feel unconcerned about group delay, that does not mean that the person does not know what he is talking about.
Some time back, the people at Stryke ran some tests and discovered that group delay for a sub is associated with transient response. The higher the group delay, the worse the transient response. So a low group delay is nice to have, if for that reason alone.
Back in the seventies, people were pooh-poohing phase relationships in speakers as a fad. Over the course of time, it has been accepted that phase is important for good reproduction. Just because something is new doesn't mean it is overrated.
I kind of agree that bass from a larger woofer sounds better. But the larger the woofer, the bigger the box for the same cutoff. So it is a tradeoff.
I am glad that the poster came forward to explain group delay. It is a term that is getting increasing attention, an I am glad that someone came forward to tell us what it is. If you don't "get" the explanation, or simply feel unconcerned about group delay, that does not mean that the person does not know what he is talking about.
Some time back, the people at Stryke ran some tests and discovered that group delay for a sub is associated with transient response. The higher the group delay, the worse the transient response. So a low group delay is nice to have, if for that reason alone.
Back in the seventies, people were pooh-poohing phase relationships in speakers as a fad. Over the course of time, it has been accepted that phase is important for good reproduction. Just because something is new doesn't mean it is overrated.
SY said:
I always like to use 2 drivers in a box push-push, 4 for a stereo pair. i have also been finding that i get better integration if the bass driver has an extended range and have therefore been heading towards older 12s (if you are patient they can be really cheap too
), or 10s or even 8s if i need to use new drivers. This often means giving up some ultimate extension on the bottom, but it isn't really missed.dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.