ok
i have to admit that i do not understand or grasps half of what you guys are talking about ..
if we summarize ...
W-Frame advantages and H-Frame advatages
inconvenient and prerequisites of the drivers for both also
Then design wise,
do the drivers need to be at 90degree perfectly for a W-Frame magic to happen ?
is there any other design rules to follow when constructing a W or an H baffle
i also didn't find the Linkwitz paper you guys are all refering to .
and both needs to be seriously braised enclosures because of obvious vibrations/resonnance?
I would like to know wich design gives the best quality wise performance ( sonically )
also wich drawbacks/solutions need to be adressed when using this design
super high spl at low hz aren't my concern with this project, but super high quality is
with 4 12 or 15" drivers in a small room i should'n't have much trouble making up for the lack of efficience in lower octaves i guess 😛
i have to admit that i do not understand or grasps half of what you guys are talking about ..
if we summarize ...
W-Frame advantages and H-Frame advatages
inconvenient and prerequisites of the drivers for both also
Then design wise,
do the drivers need to be at 90degree perfectly for a W-Frame magic to happen ?
is there any other design rules to follow when constructing a W or an H baffle
i also didn't find the Linkwitz paper you guys are all refering to .
and both needs to be seriously braised enclosures because of obvious vibrations/resonnance?
I would like to know wich design gives the best quality wise performance ( sonically )
also wich drawbacks/solutions need to be adressed when using this design
super high spl at low hz aren't my concern with this project, but super high quality is
with 4 12 or 15" drivers in a small room i should'n't have much trouble making up for the lack of efficience in lower octaves i guess 😛
That´s exactly how I started into dipoles. So your premises are perfect. 😀
I don´t think there is a special W-frame magic. Any configuration from plain flat baffle to a compact "ripole" might work. Driver requirements would be almost identical for all applications. But please don´t mix different drivers in one baffle.
www.linkwitzlab.com A MUST read (at least a big part of it) if you are seriously into designing your individual dipole frame. And you will find those design rules too.
Indeed. It was my fault to show my W-baffle without the final bracing in the middle of the front 🙁 . JohninCR is definitely right in putting some emphasis on this.
Impulse compensation: Full for 180° W-baffle a la Linkwitz, half for 90° W-baffle, none for H-baffle.
Efficiency: Best for H-baffle, worst for ripole. The more you fold the baffle, the more efficiency you will loose.
Usable upper frequency limit: Highest for H-Baffle, lowest for a sqeezed ripole
For overall sonic quality the driver quality will be more important than the baffle geometry IMHO. I believe that most geometry decicions are done for reasons regarding volume, apppearance or geometry and not for sound quality in the first place.
I hope this keeps you busy for some time. 😉
Rudolf
Open Baffle or Infinite Baffle
I don't know if you have investigated infinite baffle subwoofers, but I thought that I would at least mention them as an alternative to open baffle designs.
Infinite baffle subwoofers have similar sound qualities to open baffle subs, although they MAY have more room node issues (but I am not convinced at that), but don't suffer the gross inefficiencies that open baffle designs suffer from. An infinite baffle design uses a crawl space, an attic or a second room as the "box" for the backwave Thus there is very little backpressure on the driver and has a similar sound quality to an open baffle design, but because there is no sound cancellation from front to back, the deep bass extension is much better.
With my open baffle system (right now), I am only flat down to 60 hz (I don't have an electronic crossover), although I will get quite a bit lower when I add my rectangular box to the back of the W open baffle box. However, with an infinite baffle system, which enjoys the benefits of room gain, it is normal to achieve flat frequency response down to 20 hz, and some are even flat down to under 10 hz.
If I am able to convert a backroom, which has a crawlspace, to an audio room, I will most likely design an infinite baffle subwoofer into the room design.
Are you in the position to consider an infinite baffle subwoofer?
Retsel
I don't know if you have investigated infinite baffle subwoofers, but I thought that I would at least mention them as an alternative to open baffle designs.
Infinite baffle subwoofers have similar sound qualities to open baffle subs, although they MAY have more room node issues (but I am not convinced at that), but don't suffer the gross inefficiencies that open baffle designs suffer from. An infinite baffle design uses a crawl space, an attic or a second room as the "box" for the backwave Thus there is very little backpressure on the driver and has a similar sound quality to an open baffle design, but because there is no sound cancellation from front to back, the deep bass extension is much better.
With my open baffle system (right now), I am only flat down to 60 hz (I don't have an electronic crossover), although I will get quite a bit lower when I add my rectangular box to the back of the W open baffle box. However, with an infinite baffle system, which enjoys the benefits of room gain, it is normal to achieve flat frequency response down to 20 hz, and some are even flat down to under 10 hz.
If I am able to convert a backroom, which has a crawlspace, to an audio room, I will most likely design an infinite baffle subwoofer into the room design.
Are you in the position to consider an infinite baffle subwoofer?
Retsel
ahahha
that is funny
i've "looked at" and not completly read
linktwitz' site a few times...almost always looking at it's sub analysis without ever noticing that his double compound subwoofer was a 180 folded baffle
this is a seriously compact design though
i still wonder a few things
what exactly is :Impulse compensation
and then is this more important than mechanical distortion wich W frame permits to cancel ??
i'd guess that all those designs have a compromise over all the positive and negatives ...
and i still don't get what i a "U" frame ?
is it a sort of short TL with the driver at the end of the enclosure?
what is called the woofer arangement on linkwitz Orion ?
is this an H-frame type ?
So W is more compact and offers impulse compensation
AND mechanical cancellation ?
i'll have to think about all that and seriously READ linkwitz papers ..
thanks again Rud 🙂
that is funny
i've "looked at" and not completly read
linktwitz' site a few times...almost always looking at it's sub analysis without ever noticing that his double compound subwoofer was a 180 folded baffle
this is a seriously compact design though
i still wonder a few things
what exactly is :Impulse compensation
and then is this more important than mechanical distortion wich W frame permits to cancel ??
i'd guess that all those designs have a compromise over all the positive and negatives ...
and i still don't get what i a "U" frame ?
is it a sort of short TL with the driver at the end of the enclosure?
what is called the woofer arangement on linkwitz Orion ?
is this an H-frame type ?
So W is more compact and offers impulse compensation
AND mechanical cancellation ?
i'll have to think about all that and seriously READ linkwitz papers ..
thanks again Rud 🙂
You should also read JohnK's tech studies at www.MusicAndDesign.com
I consider anything with more baffle to the rear a U-baffle. A pure U would be a driver on the end of a short pipe with damping to eliminate the 1/4 wave resonances. If you took the Linkwitz W-dipole woofer, not subwoofer (it rolls off too early), and slid the driver manifold the 2.25" to the front, you'd pick up 4.5" of extra rear wave travel distance. This brings the Fequal point, where 6db/oct roll off starts down from 121hz in the Linkwitz cab to only 97hz with exactly the same size enclosure, and negible, if any, sonic difference. It does shift away from pure dipole, and this W/U hybrid would have a slight shift rearward in the area of greatest null. The "U" part is the rearward extension that isn't equal in the front.
Above is a perfect example of why I would never again build a pure dipole to handle the bass region. If you want the response of an H, cut the front off of it and eliminate the potential on the front side for resonances. The only time I can see any justification for any panel to extend toward the front past the driver mounting baffle is be able to hide the magnet in a push/pull alignment. There's always room on the backside of any U for damping, if necessary. Small changes can have big benefits, but change the polar response pattern very little, so room problems aren't a concern. Your net imput into the room is still 0, so your bass stays in the room very well (louder listening without disturbing others). Plus you can always get the same response in a smaller enclosure, plus you avoid the dipole bottom extension limit in small rooms. There aren't too many win-win compromises in audio, but I see using U hybrids in the bass region as one of them.
I consider anything with more baffle to the rear a U-baffle. A pure U would be a driver on the end of a short pipe with damping to eliminate the 1/4 wave resonances. If you took the Linkwitz W-dipole woofer, not subwoofer (it rolls off too early), and slid the driver manifold the 2.25" to the front, you'd pick up 4.5" of extra rear wave travel distance. This brings the Fequal point, where 6db/oct roll off starts down from 121hz in the Linkwitz cab to only 97hz with exactly the same size enclosure, and negible, if any, sonic difference. It does shift away from pure dipole, and this W/U hybrid would have a slight shift rearward in the area of greatest null. The "U" part is the rearward extension that isn't equal in the front.
Above is a perfect example of why I would never again build a pure dipole to handle the bass region. If you want the response of an H, cut the front off of it and eliminate the potential on the front side for resonances. The only time I can see any justification for any panel to extend toward the front past the driver mounting baffle is be able to hide the magnet in a push/pull alignment. There's always room on the backside of any U for damping, if necessary. Small changes can have big benefits, but change the polar response pattern very little, so room problems aren't a concern. Your net imput into the room is still 0, so your bass stays in the room very well (louder listening without disturbing others). Plus you can always get the same response in a smaller enclosure, plus you avoid the dipole bottom extension limit in small rooms. There aren't too many win-win compromises in audio, but I see using U hybrids in the bass region as one of them.
Hi,
"For overall sonic quality the driver quality will be more important than the baffle geometry IMHO"
That is soo true! imho 😀
Design-aims as optics and size may play and do play a great role in the dipole design, because technically the differences don´t make much sonic difference as long as You plan to use the thing as a subwoofer below 100Hz. But for the good feel, symmetry and small size are the sexiest.
(btw. the Linkwitz-W is rather big and ugly to my eyes and not symmetrical as a ripole). And I fear that the advantages of the distortion cancellation by mounting the drivers as SL does will be swamped by the inherent asymmetry of the construction.
jauu
Calvin
"For overall sonic quality the driver quality will be more important than the baffle geometry IMHO"
That is soo true! imho 😀
Design-aims as optics and size may play and do play a great role in the dipole design, because technically the differences don´t make much sonic difference as long as You plan to use the thing as a subwoofer below 100Hz. But for the good feel, symmetry and small size are the sexiest.
(btw. the Linkwitz-W is rather big and ugly to my eyes and not symmetrical as a ripole). And I fear that the advantages of the distortion cancellation by mounting the drivers as SL does will be swamped by the inherent asymmetry of the construction.jauu
Calvin
Uh...John, I think you might want to re-read both Siegfried's and John K's excellent technical explanations of the various woofer configurations and how the "D", "D/2", "Fequal", etc, etc, parameters are related in the three configs.
Cheers,
Davey.
Cheers,
Davey.
Davey,
If you have a point then state it. It sounds like you're the one who needs go back and re-read that great work. Most do misunderstand how D is measured and and it's impact on Fequal once you fold a baffle. This is due in part because of the way SL displays D graphically in the H configuration. In an H dipole, the distance differential is from the back of the driver baffle around the rear edge back to the front to the plane of the driver baffle. SL displays it as the side panel depth, which is only correct because they happen to be equal. John K also has the tendency to explain things in a technical manner, and as a result very simple explanations are sometimes not spelled out.
The very simple explanation of why a U-baffle has 6db greater output at the bottom than a similarly dimensioned dipole H-baffle is because D is doubled in a U configuration. Doubling D moves the Fequal point down exactly 1 octave, which is exactly a +6db increase.
The D that I am talking about is not Diameter. It is the Differential in travel distance for the rear wave vs the front wave. I believe at some point SL uses D as diameter, then later as the Differential in travel distance, which leads to confusion.
If you have a point then state it. It sounds like you're the one who needs go back and re-read that great work. Most do misunderstand how D is measured and and it's impact on Fequal once you fold a baffle. This is due in part because of the way SL displays D graphically in the H configuration. In an H dipole, the distance differential is from the back of the driver baffle around the rear edge back to the front to the plane of the driver baffle. SL displays it as the side panel depth, which is only correct because they happen to be equal. John K also has the tendency to explain things in a technical manner, and as a result very simple explanations are sometimes not spelled out.
The very simple explanation of why a U-baffle has 6db greater output at the bottom than a similarly dimensioned dipole H-baffle is because D is doubled in a U configuration. Doubling D moves the Fequal point down exactly 1 octave, which is exactly a +6db increase.
The D that I am talking about is not Diameter. It is the Differential in travel distance for the rear wave vs the front wave. I believe at some point SL uses D as diameter, then later as the Differential in travel distance, which leads to confusion.
That's some great work. It brings up some very interest questions, especially since the actual and predicted SPL tracks so closely below 150hz.
First, he predicts a rise of 12db from Fequal up to Fmax. My understanding from Linkwitz's work is the peak should be at +6db.
Second, while panel resonance and/or 1/4 wave resonance could explain the peak around 180hz, the steep null in the next octave looks like it has to be the dipole null at 2 x Fmax. What is causing the dipole peak behavior half an Octave earlier than expected? I hope it means that something about the manifold alignment makes what should be a 100hz Fequal baffle according to Linkwitz in free space, a 60hz Fequal in the real world. Maybe the rear pressure wave needs to clear the rear cavity by some distance before it starts its true free space expansion in all directions including back toward the front, because the peak at 180hz indicates almost a 1 meter separation distance vs the 60cm distance in his cab drawings. This is quite a significant difference, and may be part of why I get deeper bass than expected in folded alignments, which I always attributed to room benefit. More likely though, it is because he measured at only 2 meters, but that still doesn't explain the early Fmax.
First, he predicts a rise of 12db from Fequal up to Fmax. My understanding from Linkwitz's work is the peak should be at +6db.
Second, while panel resonance and/or 1/4 wave resonance could explain the peak around 180hz, the steep null in the next octave looks like it has to be the dipole null at 2 x Fmax. What is causing the dipole peak behavior half an Octave earlier than expected? I hope it means that something about the manifold alignment makes what should be a 100hz Fequal baffle according to Linkwitz in free space, a 60hz Fequal in the real world. Maybe the rear pressure wave needs to clear the rear cavity by some distance before it starts its true free space expansion in all directions including back toward the front, because the peak at 180hz indicates almost a 1 meter separation distance vs the 60cm distance in his cab drawings. This is quite a significant difference, and may be part of why I get deeper bass than expected in folded alignments, which I always attributed to room benefit. More likely though, it is because he measured at only 2 meters, but that still doesn't explain the early Fmax.
John,
I believe it was GM who told me that the effective length of any resonator (be it quarter wave pipe or BR port) has to include some correction factor, which should be roughly half the diameter of the tube/pipe/port. So your explanation of the lowered resonance seems logic.
I´ve seen some math on the net for calculating that correction factor for oddly shaped openings, but different approaches seem to give different results.
Rudolf
I believe it was GM who told me that the effective length of any resonator (be it quarter wave pipe or BR port) has to include some correction factor, which should be roughly half the diameter of the tube/pipe/port. So your explanation of the lowered resonance seems logic.
I´ve seen some math on the net for calculating that correction factor for oddly shaped openings, but different approaches seem to give different results.
Rudolf
Rudolph,
GM likes to use the visualization of a pressure or sound wave as a bubble travelling through a pathway, and if we visualize what is happening as a wave being a bubble, then it makes sense that the bubble can't immediately expand in the forward on axis direction with its full potential because the cab is in the way.
If this is true, then it's good and bad news for those of us who advocate manifolds and U-baffle shapes. While we can get some significant free ride in terms of "D", Differential of front/rear wave travel distance, the cost will probably be at the expense of lowering the frequencies where resonances begin.
Time to consult the physics experts. I like free rides and I address resonances with damping anyway.
GM likes to use the visualization of a pressure or sound wave as a bubble travelling through a pathway, and if we visualize what is happening as a wave being a bubble, then it makes sense that the bubble can't immediately expand in the forward on axis direction with its full potential because the cab is in the way.
If this is true, then it's good and bad news for those of us who advocate manifolds and U-baffle shapes. While we can get some significant free ride in terms of "D", Differential of front/rear wave travel distance, the cost will probably be at the expense of lowering the frequencies where resonances begin.
Time to consult the physics experts. I like free rides and I address resonances with damping anyway.
Free rides in physics? 🙂
Regarding Stefan's excellent effort... Simulations are great....to a point. But when the real-world measurements show a significant discrepancy then there's only one conclusion. The simulation is incorrect.
I believe/understood Ingemar was continuing to work on the program to try and make more accurate the "box," "tube," and "mouth" parameters. Also, it's not clear the numbers for those parameters used by Stefan are correct. You'll note that he only included half the volume and distance of the chambers in his CAD model.
It just goes to show that (in conjunction with the drivers) this is a more complicated acoustic structure than it would first appear.
I think the 180Hz peak in his measurements is indeed the cavity resonance. This is consistent with "W"-frame enclosures I've built with similar dimensions.
Cheers,
Davey.
Regarding Stefan's excellent effort... Simulations are great....to a point. But when the real-world measurements show a significant discrepancy then there's only one conclusion. The simulation is incorrect.
I believe/understood Ingemar was continuing to work on the program to try and make more accurate the "box," "tube," and "mouth" parameters. Also, it's not clear the numbers for those parameters used by Stefan are correct. You'll note that he only included half the volume and distance of the chambers in his CAD model.
It just goes to show that (in conjunction with the drivers) this is a more complicated acoustic structure than it would first appear.
I think the 180Hz peak in his measurements is indeed the cavity resonance. This is consistent with "W"-frame enclosures I've built with similar dimensions.
Cheers,
Davey.
Davey said:
Davey,
OK, how about a free ride compared to common theory due to the actual physics?
How sure are you that is wasn't panel resonance? I've had tremendously audible panel resonances with unbraced folded forms. Damping clears up the cavity resonances, with no sonic detriment to my ears, but they need to be avoided on the front side. I hope to avoid them using non-parallel surfaces and small drivers.
- Status
- Not open for further replies.