I somehow got the notion that Q=0.5 was ideal for a subwoofer, and that the more you strayed from that number, the worse the transient response would be. Now I discover speakergod Linkwitz putting a Peerless XLS 12" in a 50 liter sealed box.
http://www.woodartistry.com/linkwitzlab/thor-intro.htm
What's up with that? I calculate a Q of 0.28. Am I missing something?
It looks like I am never going to build my first subwoofer, because I am in a perpetual state of confusion. How does a person ever learn this stuff????
http://www.woodartistry.com/linkwitzlab/thor-intro.htm
What's up with that? I calculate a Q of 0.28. Am I missing something?
It looks like I am never going to build my first subwoofer, because I am in a perpetual state of confusion. How does a person ever learn this stuff????
Speaker Qt
Generally, drivers with a Qts of .2 up to .5 will work well in a reflex enclosure.
Sealed boxes work best in the higher end of this range, say from .35 up to .55 Qts.
Open or infinite baffle speakers are .6 and above.
When you see a driver Qts in the mid to high .30's, it will usually work well in either enclosure.
If you think about Qts as a reflection of motor strength vs mechanical properties like compliance, the higher the number, the less damping of the cone; lower number means stronger motor and greater cone damping.
Loosely speaking, a sealed box has more cone movement than a reflex box, and requires slightly less cone control, or higher Qts.
Although, as a driver passes resonance (Fs) in a reflex box, all kinds of movement can occur, and the particular Q of the sealed alignment can alter cone movement, too.
Select a driver with a Qts of .45, for example, and winISD will default to a sealed enclosure, if you click "next" and don't choose a box type. The program is picking the enclosure best suited to the TS parameters for you.
The small bar graph in the "box" tab is a visual indicator of the flexibility of the driver in different alignments, with various color bands showing up.
Tim
Generally, drivers with a Qts of .2 up to .5 will work well in a reflex enclosure.
Sealed boxes work best in the higher end of this range, say from .35 up to .55 Qts.
Open or infinite baffle speakers are .6 and above.
When you see a driver Qts in the mid to high .30's, it will usually work well in either enclosure.
If you think about Qts as a reflection of motor strength vs mechanical properties like compliance, the higher the number, the less damping of the cone; lower number means stronger motor and greater cone damping.
Loosely speaking, a sealed box has more cone movement than a reflex box, and requires slightly less cone control, or higher Qts.
Although, as a driver passes resonance (Fs) in a reflex box, all kinds of movement can occur, and the particular Q of the sealed alignment can alter cone movement, too.
Select a driver with a Qts of .45, for example, and winISD will default to a sealed enclosure, if you click "next" and don't choose a box type. The program is picking the enclosure best suited to the TS parameters for you.
The small bar graph in the "box" tab is a visual indicator of the flexibility of the driver in different alignments, with various color bands showing up.
Tim
Okay, I think I understand how EQ affects resonance. What we are interested in (I guess) is the ratio of stored energy to energy going into the equalizer (not the speaker). Right? In other words, the Q of interest is that of the system that comprises the speaker, box, and equalizer.
So now I'm thinking about just getting a 1 cu ft box and slapping a Peerless XLS 10" in there with a plate amp that has a 6db boost at 30Hz. With the room gain in my small office, that should do quite well, I think, maybe, perhaps, unless I am mistaken.
Here's a graph of the speaker in a nominal 26 liter box and no EQ. Whatcha think?
So now I'm thinking about just getting a 1 cu ft box and slapping a Peerless XLS 10" in there with a plate amp that has a 6db boost at 30Hz. With the room gain in my small office, that should do quite well, I think, maybe, perhaps, unless I am mistaken.
Here's a graph of the speaker in a nominal 26 liter box and no EQ. Whatcha think?
Attachments
Suitable woofer
Just look at the properties you need for the ideal woofer for the task, for a minute.
That 1 cu. ft. enclosure size does not give you any low frequency extension to speak of, with the woofer starting to roll off at 300 Hz, and being well down at 30 Hz. With a low Qts woofer in a sealed box, your chances for success are limited. Use a woofer near .4 Qts or higher for a sealed enclosure.
Sealed box:
Room gain in a perfectly sealed room is 12 dB/octave - so 9 dB is more like it in the real world. If you match the slope of the roll off to 9 dB gain, you will see very low extension. Right now your slope is more like a reflex enclosure and too steep for room lift to help.
Your box size is not permitting this, indicating too much woofer for the box, or a poor choice for a small sealed sub.
Tim
Just look at the properties you need for the ideal woofer for the task, for a minute.
That 1 cu. ft. enclosure size does not give you any low frequency extension to speak of, with the woofer starting to roll off at 300 Hz, and being well down at 30 Hz. With a low Qts woofer in a sealed box, your chances for success are limited. Use a woofer near .4 Qts or higher for a sealed enclosure.
Sealed box:
Room gain in a perfectly sealed room is 12 dB/octave - so 9 dB is more like it in the real world. If you match the slope of the roll off to 9 dB gain, you will see very low extension. Right now your slope is more like a reflex enclosure and too steep for room lift to help.
Your box size is not permitting this, indicating too much woofer for the box, or a poor choice for a small sealed sub.
Tim
Re: Suitable woofer
9dB/Octave seems like too large a number for room gain for the 20Hz to 80Hz range. Where did your number come from?
Adire Audio uses a room gain of about 4 dB from 80Hz to 20Hz in their subwoofer design white papers. This is only about 2dB/octave.
They show the overall room gain from 20Hz to 300Hz to be about 12dB overall but the gain in the 20Hz to 80Hz range is somewhat shallow.
It is not clear what size room they are using for a "typical" room in their model. Presumably something like a 4m x 4m room. Something like a car would have a higher gain that starts at a higher frequency.
Tim Moorman said:
9dB/Octave seems like too large a number for room gain for the 20Hz to 80Hz range. Where did your number come from?
Adire Audio uses a room gain of about 4 dB from 80Hz to 20Hz in their subwoofer design white papers. This is only about 2dB/octave.
They show the overall room gain from 20Hz to 300Hz to be about 12dB overall but the gain in the 20Hz to 80Hz range is somewhat shallow.
It is not clear what size room they are using for a "typical" room in their model. Presumably something like a 4m x 4m room. Something like a car would have a higher gain that starts at a higher frequency.
Remember the impedance curve!
Some arguments in favour of a low Qts woofer in a small sealed box:
1) In most cases of a sealed woofer, the impedance peaks at a comfortably high value somewhere around 30Hz~70Hz, right where electrical equalization is most needed. This means that the woofer can be equalized to perform similarly to a ported or high Qts woofer, but with a cooler amp and smaller box.
2) A low Qts means that the back-pressure behind the speaker is blocked more effectively than with a high Qts woofer. This is because the stronger magnet allows the amplifier to have better control over the voice coil, so it can resist external forces more effectively. It may appear that a high Qts woofer would have an advantage because the mechanical resonance produces a useful bass boost, however: the 30~70Hz resonance is not the only resonance that is produced. Inevitably there are also internal box colourations, and they may be audible especially if a gentle crossover slope is desired. A low Qts speaker will block these colourations much more effectively, and the differences in low-frequency response show just how important the improvement could be at higher frequencies.
3) A sealed air spring is much more linear than a speaker's spider suspension, and the air suspension will be the dominant factor in a small sealed box especially if the VAS of the speaker is high. This will result in lower excursion-related distortion.
CM
Some arguments in favour of a low Qts woofer in a small sealed box:
1) In most cases of a sealed woofer, the impedance peaks at a comfortably high value somewhere around 30Hz~70Hz, right where electrical equalization is most needed. This means that the woofer can be equalized to perform similarly to a ported or high Qts woofer, but with a cooler amp and smaller box.
2) A low Qts means that the back-pressure behind the speaker is blocked more effectively than with a high Qts woofer. This is because the stronger magnet allows the amplifier to have better control over the voice coil, so it can resist external forces more effectively. It may appear that a high Qts woofer would have an advantage because the mechanical resonance produces a useful bass boost, however: the 30~70Hz resonance is not the only resonance that is produced. Inevitably there are also internal box colourations, and they may be audible especially if a gentle crossover slope is desired. A low Qts speaker will block these colourations much more effectively, and the differences in low-frequency response show just how important the improvement could be at higher frequencies.
3) A sealed air spring is much more linear than a speaker's spider suspension, and the air suspension will be the dominant factor in a small sealed box especially if the VAS of the speaker is high. This will result in lower excursion-related distortion.
CM
XLS Sealed
Ceramic Man
If you run this through a box program like winISD, the alignment needed to get an acceptable qtc of .7 (ideally damped) is a 5.4 liter box, about 10 l for a highly damped qtc of .55. Anything larger is going to sound bad with qtc in the way overdamped range...
I'm all for sealed boxes, even with low Qts drivers in them, but they still must yield acceptable alignments with qtc of, say, .55 - .8 range.
Put this woofer in a vented enclosure like it was designed for, and you can contour the slopes and group delay to get fairly close to the same target, with better response.
Tim
Ceramic Man
If you run this through a box program like winISD, the alignment needed to get an acceptable qtc of .7 (ideally damped) is a 5.4 liter box, about 10 l for a highly damped qtc of .55. Anything larger is going to sound bad with qtc in the way overdamped range...
I'm all for sealed boxes, even with low Qts drivers in them, but they still must yield acceptable alignments with qtc of, say, .55 - .8 range.
Put this woofer in a vented enclosure like it was designed for, and you can contour the slopes and group delay to get fairly close to the same target, with better response.
Tim
Re: XLS Sealed
Have you read the whole thread? What started this off was my observation that Sigfried Linkwitz designed a sealed box sub that had a Q of .4 or so, unequalized. Presumably he knew what he was doing.
Also, there is considerable room for discussion about which Q is "ideal". How about 0.5? Anyone for 0.577?
Tim Moorman said:
Have you read the whole thread? What started this off was my observation that Sigfried Linkwitz designed a sealed box sub that had a Q of .4 or so, unequalized. Presumably he knew what he was doing.
Also, there is considerable room for discussion about which Q is "ideal". How about 0.5? Anyone for 0.577?
low q is good!!
Low speaker Q is good! I imagine speaker Q is a misunderstood parameter among most audio enthusiasts. I did not understand the meaning of speaker Q until after I went through engineering school. I have been an audio hobbyist for over twenty years and have a bachelor of science degree in electrical engineering. I don't want to sound arrogant but with a degree in electrical engineering just trust I know what I am talking about and ignore any advice from amatuers advocating anything other than a low Q. So, what in the freakin' hell is speaker Q anyway? To understand Q one must understand the concept of speaker resonant frequency. Think of a mass suspended by a coil spring. Tug on the mass suspended from a spring and it will bounce up and down at a certain 'frequency'. A larger mass results in a lower resonant frequency. A smaller mass results in a higher frequency. A stiffer spring results in a higher frequency. A more less stiff (more compliant) spring results in a lower frequency. This is exactly the way a woofer ends up with a resonant frequency. A massive cone and low stiffness suspension will result in a low resonant frequency. A low mass, high stiffness suspension will result in a high resonant frequency. To reproduce low frequency sound (bass) efficiently it is desireable to have a low resonant frequency. There are two ways to end up with a low resonant frequency. A high mass speaker cone in conjunction with a relatively stiff suspension is one way. A low mass cone with a low stiffness (high compliance) suspension is another way to end up with a low frequency of resonance. But mass and stiffness are not the only parameters involved. Damping is the third parameter which helps determine speaker Q. Damping is akin to friction. Friction like a 'shock absober' on your car. The shock absorber on your car does nothing but add friction to keep your car from bouncing up and down and up and down forever. Friction is necessary to keep your car from bouncing too much. And yes there is friction in a speaker too. The lower this friction is the more pronounced the peak of cone resonance is. In other words if you were to plot a graph of the speaker resonance on a frequency scale a speaker with low damping would peak very high at it's resonant frequency. If you were to compare different speakers and graph their resonance a speaker with a narrow high peak (less damping) at it's resonant frequency would be a high Q speaker. If the curve is broader and flatter (more damping) then this is a low Q speaker. If this all sounds confusing bear with me I'm getting to the good part. Unlike the suspension on your car it is not desireable to have a lot of friction restricting the movement of the speaker cone but some friction is inevitable. It's the RATIO of this inherent friction to the mass and stiffness of the speaker which determines speaker Q. If two speakers have the SAME amount of frictional losses the low mass, low stiffness speaker will have a lower Q and a high mass, high stiffness speaker will have a high Q. A low mass cone, low stiffness speaker has one major advantage over a high mass cone. The low mass cone can respond much more quickly to transients. In other words, a low mass cone will be much much more crisp and 'punchy' when transients like drum hits are being reproduced. Low Q speakers have MUCH better transient response. Adjectives like 'tight' have been used to describe a good transient response. Adjectives like 'muddy' and 'boomy' have been used to describe high Q speakers with lousy transient response. In order to make a lightweight cone stiff one must employ high dollar materials such as Kevlar and carbon fiber. High quality low Q speakers will tend to be more expensive. Is there any time a high Q speaker is better? Yes, if you only care about reproducing low frequenct sound at high SPL levels for some retarded sound contest and you care nothing about high fidelity then high Q is what you want. By the way, the peerless XLS are excellent examples of good low Q speakers. I used these in a system I built for a friend. The transient response and bass output are AWESOME!!!
Jim
Low speaker Q is good! I imagine speaker Q is a misunderstood parameter among most audio enthusiasts. I did not understand the meaning of speaker Q until after I went through engineering school. I have been an audio hobbyist for over twenty years and have a bachelor of science degree in electrical engineering. I don't want to sound arrogant but with a degree in electrical engineering just trust I know what I am talking about and ignore any advice from amatuers advocating anything other than a low Q. So, what in the freakin' hell is speaker Q anyway? To understand Q one must understand the concept of speaker resonant frequency. Think of a mass suspended by a coil spring. Tug on the mass suspended from a spring and it will bounce up and down at a certain 'frequency'. A larger mass results in a lower resonant frequency. A smaller mass results in a higher frequency. A stiffer spring results in a higher frequency. A more less stiff (more compliant) spring results in a lower frequency. This is exactly the way a woofer ends up with a resonant frequency. A massive cone and low stiffness suspension will result in a low resonant frequency. A low mass, high stiffness suspension will result in a high resonant frequency. To reproduce low frequency sound (bass) efficiently it is desireable to have a low resonant frequency. There are two ways to end up with a low resonant frequency. A high mass speaker cone in conjunction with a relatively stiff suspension is one way. A low mass cone with a low stiffness (high compliance) suspension is another way to end up with a low frequency of resonance. But mass and stiffness are not the only parameters involved. Damping is the third parameter which helps determine speaker Q. Damping is akin to friction. Friction like a 'shock absober' on your car. The shock absorber on your car does nothing but add friction to keep your car from bouncing up and down and up and down forever. Friction is necessary to keep your car from bouncing too much. And yes there is friction in a speaker too. The lower this friction is the more pronounced the peak of cone resonance is. In other words if you were to plot a graph of the speaker resonance on a frequency scale a speaker with low damping would peak very high at it's resonant frequency. If you were to compare different speakers and graph their resonance a speaker with a narrow high peak (less damping) at it's resonant frequency would be a high Q speaker. If the curve is broader and flatter (more damping) then this is a low Q speaker. If this all sounds confusing bear with me I'm getting to the good part. Unlike the suspension on your car it is not desireable to have a lot of friction restricting the movement of the speaker cone but some friction is inevitable. It's the RATIO of this inherent friction to the mass and stiffness of the speaker which determines speaker Q. If two speakers have the SAME amount of frictional losses the low mass, low stiffness speaker will have a lower Q and a high mass, high stiffness speaker will have a high Q. A low mass cone, low stiffness speaker has one major advantage over a high mass cone. The low mass cone can respond much more quickly to transients. In other words, a low mass cone will be much much more crisp and 'punchy' when transients like drum hits are being reproduced. Low Q speakers have MUCH better transient response. Adjectives like 'tight' have been used to describe a good transient response. Adjectives like 'muddy' and 'boomy' have been used to describe high Q speakers with lousy transient response. In order to make a lightweight cone stiff one must employ high dollar materials such as Kevlar and carbon fiber. High quality low Q speakers will tend to be more expensive. Is there any time a high Q speaker is better? Yes, if you only care about reproducing low frequenct sound at high SPL levels for some retarded sound contest and you care nothing about high fidelity then high Q is what you want. By the way, the peerless XLS are excellent examples of good low Q speakers. I used these in a system I built for a friend. The transient response and bass output are AWESOME!!!
Jim
But you are only talking about Qm here. Don't forget that Qt is (Qm*Qe)/(Qm + Qe).
The electrical Q is a function of the strength of the field in the gap, and the resistance of the voice coil. You can manipulate Qe to your hearts content using techniques like positive feedback of the sensed coil current to give a negative output impedance for the amplifier. You can then apply your frequency correction network.
BTW, I also prefer the smooth LF roll-of of a low-Q sealed-box system, but there's more than one way of getting this desired outcome.
The electrical Q is a function of the strength of the field in the gap, and the resistance of the voice coil. You can manipulate Qe to your hearts content using techniques like positive feedback of the sensed coil current to give a negative output impedance for the amplifier. You can then apply your frequency correction network.
BTW, I also prefer the smooth LF roll-of of a low-Q sealed-box system, but there's more than one way of getting this desired outcome.
Ouroboros said:
Not necessarily. Qes is actually the dominant factor here, and the amplifier acts like a near short circuit (resistance that's equal to the amp's output resistance) across the speaker's terminals. A Qes of 0.35 is theoretically equivalent (in terms of damping only) to a Qms of 0.35. Current flowing across the voice coil's resistance is that friction that jimseko talked about.
A hard cone material usually isn't enough for pistonic motion. Its design has to be optimized by eg: curving the cone, making it a very thick sandwich of materials (Eton), or using a very stiff material (Seas, Accuton, Visaton, Alcone and others).
It would seem that getting a punchy sounding speaker is a vicious circle, especially when recording artists are involved in any way. A person mixing a recording in a studio will mix it to sound good on his set of punchy-sounding speakers, and anyone listening to that recording will need an equally punchy-sounding pair for the same effect. It would be a bit like always having a 50Hz to 120Hz bass boost without knowing it.
That statement assumed that only the frequency response affects the low frequency sound quality, and I think that that's a big mistake that people here are making. Another important factor is the Group Delay. It indicates if there are low frequency resonances, due to the box design. For optimal "punchiness" the group delay should only have a flat or negative slope, ie: the higher the frequency - the smaller the delay, and this will ensure that there will be no bass ringing.
Bass ringing, AKA: slow, sloppy, boomy, overhung bass, can be eliminated by experimenting with different box simulations and port tunings to either make the box undersized and tuned too high (albeit with an amplitude boost), or oversized and tuned so low that the ringing is at an inaudible frequency. As a compromise, the delay peak of a ported woofer can be brought to a low frequency that is still audible, but far below the cut-off frequency so that it's not obtrusive. The third option seems to be a popular choice around here, even though some people may not realise why it sounds good.
Linkwitz's calculations can fix a problematic woofer in this sense, because they produce louder, deeper bass without wrecking the group delay curve.
Lech
More about Q
Yes, total Q is a combination of mechanical and electrical Q but I was trying to keep the explanation (relatively) simple. My main point is ALL OTHER THINGS BEING EQUAL (including the enclosure configuration) low cone mass results in better transient response and low cone mass also results in a lower Q therefore (in general) a low Q driver has a better transient response. Having a good transient response is equivalent to having a good high frequency range. (I wonder if I'll get any arguments on that one?) Too many (especially those targeted for automotive applications) woofers have a rediculously high speaker Q. The output impedance (inverse of damping factor) of the amplifer does affect total sytem Q but virtually all modern amplifiers have very low output impedance and a decent amp should have no effect on total Q. I love Rockford Fosgate amplifiers but their extremely high Q woofers suck! I've owned a pair of Rockford Fosgate 12" woofers and they sound terrible! I'd trade a Rockford Fosgate or piece-of-crap Kicker woofer for a Peerless or Adire Audio woofer in a heartbeat. I've had my eye on the Adire Maelstom 18" woofer with a Qts of .28 for a long time. Unfortunately, I would have a hard time convincing my wife I need to spend $300 on a woofer when I have already spent too much (in her opinion) on my beloved hobby! If anyone knows of a woofer with better specs than the Adire Maelstrom I'd love to know about it!!
Jim
Yes, total Q is a combination of mechanical and electrical Q but I was trying to keep the explanation (relatively) simple. My main point is ALL OTHER THINGS BEING EQUAL (including the enclosure configuration) low cone mass results in better transient response and low cone mass also results in a lower Q therefore (in general) a low Q driver has a better transient response. Having a good transient response is equivalent to having a good high frequency range. (I wonder if I'll get any arguments on that one?) Too many (especially those targeted for automotive applications) woofers have a rediculously high speaker Q. The output impedance (inverse of damping factor) of the amplifer does affect total sytem Q but virtually all modern amplifiers have very low output impedance and a decent amp should have no effect on total Q. I love Rockford Fosgate amplifiers but their extremely high Q woofers suck! I've owned a pair of Rockford Fosgate 12" woofers and they sound terrible! I'd trade a Rockford Fosgate or piece-of-crap Kicker woofer for a Peerless or Adire Audio woofer in a heartbeat. I've had my eye on the Adire Maelstom 18" woofer with a Qts of .28 for a long time. Unfortunately, I would have a hard time convincing my wife I need to spend $300 on a woofer when I have already spent too much (in her opinion) on my beloved hobby! If anyone knows of a woofer with better specs than the Adire Maelstrom I'd love to know about it!!
Jim
jimseko said:
This statement is not factual. The enclosure and Le play far more significant roles in determining transient response.
First, I use high Q drivers in dipole configuration and achieve more dynamic, detailed bass than I've ever heard from any sealed or vented woofer regardless of its driver's Qts.
Second,
I have 2 15" woofers, an Adire Tempest and a JBL 1500gti, that I have used in the same sealed enclosure. The JBL is the clear winner in terms of dynamic bass, yet it's Mms is 261 grams compared to 195 grams and it's Qts is .48 compared to .38 for the Tempest. It is heavier and has a higher Qts, but it has much better transient response. BTW, their BL's, motor force factors, are quite comparable at 15.39 and 14.2 respectively.
The big difference in Le, voice coil inductance, is what makes the JBL more dynamic. Their Le's are 1.06mH (JBL) and 2.9mH (Tempest). This voice coil inductance, in turn, is a measure of a driver's resistance to change in current which affects it's ability to respond quickly and change direction. This has a huge impact on transient response. Check out the following paper regarding woofer speed: http://www.adireaudio.com/tech_papers/woofer_speed.htm
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.