I'm quite happy for you to make a counter argument offering some actual figures, but I'm not seeing any.
I've build subs and horns before and the rear pressure wave for a sealed sub, which is what I was talking about in the previous post is not that much. All the driver does on the reverse (inward) stroke is increase the pressure in the box by the volume of air the cone displaces. Now for some of the largest displacement drivers such as the now gone Maelstrom X2, it was 34mm of one way stroke and a cone area of 1183cm2 for a total volume of 4022cc. Now for a sealed empty Qt=0.707 alignment the net volume is 123litres.
Pi.Vi = Pf.Vf
∴ Pf=Pi.Vi/Vf
So, Pf = 101.3 x 123/119 = 104.7kPa
3.4kPa = 0.49psi
Same driver in an isobaric configuration requires a rear enclosure volume of 65.6 litres.
101.3 x 65.6/61.6 = 107.88kPa
though why you would submit to such as waste of the inner driver and all that power for a bric config, I'll never know.
6.58kPa = 0.95psi
Now for a properly braced sub enclosure made out of say 18mm ply or even mdf, the displacement in the panels is going to be very low. The enclosure does not have to withstand much pressure difference at all, less than a psi in an extreme case, of one of the highest displacement subs made used wastefully. Even the LMS5400 has a bit less.
One of the reasons I mentioned my 15" drivers in the previous post was very few people even have drivers capable of that much displacement as subs, let alone MB.
I'm happy for you to show how much panel displacement will be caused by a PSI difference, for the extreme case, in a decently braced matrix or semi matrix enclosure made of the most commonly used materials to construct speakers. Of course if you're using cardboard, that's different.
Building the 123L net MalX sealed box above using ply/mdf and a 200-300mm C-C spacing for the braces in almost all realistic form factors will be very unlikely to have a panel resonance anywhere near the passband.
I don't think you have said what drivers, enclosure volumes and alignment you're using, but if you post I'll happily run the numbers again. It would be interesting to know anyway.
Brett, LOL we're trying to school each other on-list but we agree on the numbers and probably everything else. We both agree the pressure doubles for an equivalent-Q isobaric. I estimated 1 PSI, you said .95 for a larger box (like for an engine, compression involves displacement and chamber volume). Actually, that's for half the displacement, and there's an equal opposite pressure decrease, so the peak to peak pressure difference is twice that. Damn close, and yes it's not a lot of PSI.
The only place we disagree is that you trivialized the primary and most difficult function of a sealed box, saying this common cause of significant problems is "not a big deal". The application of a few psi across a large number of square inches becomes significant. Only a very small movement of a large panel makes a lot of output. Controlling panel movement so the backwave is absorbed and not radiated is the most important job of a sealed box, and a big difference between the 'sounds' of different sealed boxes of the same volume and stuffing. When the box dimensions are all less than a wavelength, you can forget about wavefronts and reflections for the most part, and focus on pressurization / depressurization forces which try to expand and contract the entire box, radiating sound that's out of phase with the drivers' real output. Since the panels move and radiate most efficiently at ther resonances, the result is a mess as the sounds combine in your room and some frequencies cancel. Very very near the box surfaces you hear its panel resonances, farther away their effect is harder to identify as those resonances disappear when mixed with the main output, leaving the effect of dropouts in the response curve. It doesn't sound like the resonance of the panels when you knock on them with your knuckles, in the room it sounds like everything EXCEPT those panel-resonant frequencies. But the forces on opposite panels are in opposite directions, so connecting the opposite panels is very effective at cancelling the forces. Make no mistake, braces prevent small panel movements, and pressurization / depressurization forces are the cause of those panel movements. Those small pressures are the bass, and are the reason we brace.
Now you qualify your trivialization saying "in a decently braced matrix or semi matrix enclosure made of the most commonly used materials..." and I agree: yes, most speakers enclosures work because they are designed to do the job.
As far as matrix bracing I love it and it works great for me. But having relevant experience I wanted to share the possible pitfalls, especially when using thin stiff matrix brace panels. Making a field of identiacal structures can make a field of sympathetic resonances, "super-tuned" with a long decay, until you damp the panels or stuff the cavities between them. And if you attach oppsite walls with thin sheet braces instead of rods or sticks etc. then if the thin sheets vibrate their sound will be magnified with in-phase output to the room by the opposed panels acting as sound boards. Holes in the matrix brace panels are good things so that internal pressure differences are not allowed to exist across the panels and cause them to move. And there are construction issues to address in making matrix bracing. The slotted sheets ideally need to be adhered to the sheets inserted, so that it makes a somewhat strong solid assembly. The attachment of the sheet braces to the cabinet walls is critical. Anchoring, bracing, or damping any free ends of panels will raise their resonant mode and can make improvements. Instead of regular or random panel spacing, consider spacing the brace panels closer near the center of the cabinet.
Best regards.
The only place we disagree is that you trivialized the primary and most difficult function of a sealed box, saying this common cause of significant problems is "not a big deal". The application of a few psi across a large number of square inches becomes significant. Only a very small movement of a large panel makes a lot of output. Controlling panel movement so the backwave is absorbed and not radiated is the most important job of a sealed box, and a big difference between the 'sounds' of different sealed boxes of the same volume and stuffing. When the box dimensions are all less than a wavelength, you can forget about wavefronts and reflections for the most part, and focus on pressurization / depressurization forces which try to expand and contract the entire box, radiating sound that's out of phase with the drivers' real output. Since the panels move and radiate most efficiently at ther resonances, the result is a mess as the sounds combine in your room and some frequencies cancel. Very very near the box surfaces you hear its panel resonances, farther away their effect is harder to identify as those resonances disappear when mixed with the main output, leaving the effect of dropouts in the response curve. It doesn't sound like the resonance of the panels when you knock on them with your knuckles, in the room it sounds like everything EXCEPT those panel-resonant frequencies. But the forces on opposite panels are in opposite directions, so connecting the opposite panels is very effective at cancelling the forces. Make no mistake, braces prevent small panel movements, and pressurization / depressurization forces are the cause of those panel movements. Those small pressures are the bass, and are the reason we brace.
Now you qualify your trivialization saying "in a decently braced matrix or semi matrix enclosure made of the most commonly used materials..." and I agree: yes, most speakers enclosures work because they are designed to do the job.
As far as matrix bracing I love it and it works great for me. But having relevant experience I wanted to share the possible pitfalls, especially when using thin stiff matrix brace panels. Making a field of identiacal structures can make a field of sympathetic resonances, "super-tuned" with a long decay, until you damp the panels or stuff the cavities between them. And if you attach oppsite walls with thin sheet braces instead of rods or sticks etc. then if the thin sheets vibrate their sound will be magnified with in-phase output to the room by the opposed panels acting as sound boards. Holes in the matrix brace panels are good things so that internal pressure differences are not allowed to exist across the panels and cause them to move. And there are construction issues to address in making matrix bracing. The slotted sheets ideally need to be adhered to the sheets inserted, so that it makes a somewhat strong solid assembly. The attachment of the sheet braces to the cabinet walls is critical. Anchoring, bracing, or damping any free ends of panels will raise their resonant mode and can make improvements. Instead of regular or random panel spacing, consider spacing the brace panels closer near the center of the cabinet.
Best regards.
Last edited:
Back to the subject of bracing and damping, what do you guys think of this stuff?
The last few days I've been brushing 5 coats of this liquid latex rubber onto the inside of the current batch of 4 cabinets, on interior walls and even braces etc. It goes on very bright pink and dries to very deep red.
RedGard 1 Gal. Waterproofing and Crack Prevention Membrane-LQWAF1 at The Home Depot
I don't know whether it will actually make much difference beyond the addition of weight lowering resonances. I don't really know whether the rubber will damp or just rebound. In the past it helped damp metal braces, and it helped damp tempered masonite matrix braces, but on these speakers the center brace is 1" thick MDF and the others are 3/4 plywood. I just put it on because the cabinet walls are thin and I had the RedGard laying around (though I ended up buying more). I used about 2 gallons on 4 cabinets.
It would be interesting to use something like this between plys or to throw in even wood scraps as braces, to constrain the rubber layer between stiffer materials. But it remains to be seen whether it makes any difference. Unfortunately, I don't have any "control" cabinet without the rubber, no time for a meaningful experiment so I won't really have any good data to share. But it is an interesting material and the price is not too outrageous as it's intended to be brushed onto concrete floors.
The last few days I've been brushing 5 coats of this liquid latex rubber onto the inside of the current batch of 4 cabinets, on interior walls and even braces etc. It goes on very bright pink and dries to very deep red.
RedGard 1 Gal. Waterproofing and Crack Prevention Membrane-LQWAF1 at The Home Depot
I don't know whether it will actually make much difference beyond the addition of weight lowering resonances. I don't really know whether the rubber will damp or just rebound. In the past it helped damp metal braces, and it helped damp tempered masonite matrix braces, but on these speakers the center brace is 1" thick MDF and the others are 3/4 plywood. I just put it on because the cabinet walls are thin and I had the RedGard laying around (though I ended up buying more). I used about 2 gallons on 4 cabinets.
It would be interesting to use something like this between plys or to throw in even wood scraps as braces, to constrain the rubber layer between stiffer materials. But it remains to be seen whether it makes any difference. Unfortunately, I don't have any "control" cabinet without the rubber, no time for a meaningful experiment so I won't really have any good data to share. But it is an interesting material and the price is not too outrageous as it's intended to be brushed onto concrete floors.
Lowering resonances is right. The weight lowers the frequency of the resonance, making them more likely to get excited and lowers the Q of the resonance, making them more audiable.
dave
It does diminish slightly the sound that results from rapping on the cabinet with your knuckles. But that's not really a very good test of how a cabinet reacts to internal pressure changes. It ends up a hard plastic rubber when dry. It seems to damp resonances a little. But I have no way of measuring how much energy is converted to heat versus just stored for rebound. I painted the interior first with that Duracoat stuff they put on commercial cabinets. Since both Duracoat and RedGard are water-based, some of the black duracoat color came up in the first coat of RedGard, and around the edges of the cabinet openings where inside meets outside the RedGard took of some of the Duracoat. The theory of constrained damping material also depends on the losses in bending the damping material. I'm just hoping this stuff is lossy (not lousy).
Okay to further push this idea: Has anyone used (Phi) ration 1.61803399 to place the Nodes in a non-resonate divisions through bracing Diagonally? In my next project I will try this method! I hope to divide the remaining LARGER panel space again into this same ration in hopes of placing any stubborn resonances at an acceptable harmonic! I would not leave out Damp-ing the braces and consequent panels with some sort of non-toxic binder like (Flex Seal + sand). Added acoustic fiber materials to take care of the mid-range/upper bass resonances also!
The knickle test is, indeed often wildly mis-interpreted.
Note that less than a 1/3 of the energy the cabinet has to deal with is from internal pressure changes, much less as the frequencies in question rise.
dave
I'm talking about stuffed sealed low-frequency boxes. Where did you come up with that? You're the second one to claim the internal pressure changes aren't a significant source of the energy the cabinet has to deal with, without any explanation. I maintain that's the primary source of the problematic forces that cause panel vibrations.
So, what are the sources of the other 2/3 of the energy? Assuming a sturdy bafleboard, once the cabinet is heavy and sturdy enough to not bounce around or deform from counterforces generated in opposition to the driver's moving mass and the energy propagated from the driver into the room, those forces are insignificant. They can tend to make a light cabinet rattle on the floor if it has crummy feet. But it's not a big contributor to panel vibrations of the cabinet's outer walls. When the cabinet dimensions are smaller than the wavelength, then internals shape, reflections, etc. are insignificant and the bass backwave is just "pressure changes" pretty uniform across the cabinet's interior area, which is why it is so effective to tie opposite walls together.
So what's the other 2/3 force source you claim? That the cabinet has to withstand big Aunt Bertha leaning on it?
Note that the higher the panel resonances the better. The diagonal brace is not as effective as the vertical brace, but i believe that a "slanted" brace that creates 2 unequal trapezoidal sub-panels would.
I have used golden ratio spacing on braces. Can't say whether it was better, certainly one does not want to place a brace such that it creates 2 equal subpanels.
dave
Attachments
I know Carver and other mfgrs who make opposed-driver 'balanced' or 'lever-balanced' speakers really like that the cabinet doesn't shake nearly as much. The whole box bounces around less, but the panel vibrations are not signficantly lessened. It is true the 'equal and opposite' reaction to the driver's moving mass and reaction to the energy propagated by the driver into the room are large forces. But they are well-coupled into the entire cabinet. To affect the panels by shaking the entire cabinet, those forces would have to shake the whole cabinet so hard the panels were flapping. The pressure changes act much more directly on the panels. Force is a vector, and the direction is very important. The problem areas are where the pressure changes act directly on the panels at 90 degrees.
I did not say that internal pressures were not significant, but that they are a maximum of 1/3 the box excitation energy.
It is trivial to prove that the last statement is false.
The cone moves. 1/2 the energy from the cone is transmitted into the box, 1/2 to the outside world (where we find it useful). An amount of energy equal to both halves is transmitted theu the basket to the cabinet. Discard the useful energy, we have 2x the energy mechanicalyl coupled to the box as the maximum possible transmitted thru the internal air cavity. The last is quite "squishy" with weakening coupling as frequency rises and losses due to damping. Hence energy tranmitted thru the air space to the box is less than a 1/3 of the total.
There is your mistake, they are not insignificant... over 2/3 of the energy transmitted into the box is via this route.
Heavy & sturdy also apply to the internal transmission so they cancel when one is comparing relative energy into the box.
dave
OK I see where you're coming from, and thanks for making me think. You're right about the 1/3 - 2/3. But that 2/3 is reasonably easy to manage with for subs with typical box shapes whereas the 1/3 due to pressure changes is a real problem.
The reaction to the driver's moving mass is 100%, and that is coupled into the basket and baffle and cabinet. However, the driver's moving mass is small (we usually wish it were smaller) compared to the mass of the basket, magnet, baffleboard, and the rest of the cabinet. Carver seems obsessed with cancelling this. Other manufacturers balance those inertial reactons with via two drivers on opposite sides or even three on the sides of a triangular enclosure. And yes the entire cabinet shakes less, but panel vibration issues remain the same.
But talking about the sound energy and opposite reaction to its propagation, which is probably what I'm guilty of underestimating: Whatever the low conversion efficiency is, like you say: a force we'll arbitrarily quantify as .5 is propagated into the room as useful output, and another .5 causes pressure changes in the cabinet which stuffing and trapped air must absorb or return as spring force. So yes, I'll agree these forces are "out of phase" and their directions add rather than cancel, and the basket/baffle/cabinet strucutre gets the counterforce of both, in twice the quantity 2*.5 = 1. That means that yes, the pressure change force is 1/2 as great as the forces from countering propagated sound, which means the air pressure changes are 1/3 th total force.
Looking at the vector direction of those forces and assuming large wavelengths and smaller box dimensions, when the driver is retracting into the box the basket is pushed in the opposite direction, pulling the box forward. The box is shaking the rear panel by its edges, and as it is moving forward the air pressure on it is increasing, the forces all additive in trying to bend the panel. And these box-moving forces are just as directional as the moving-mass reactions. So maybe the people making subs with drivers on opposite sides of the box or all around the box deserve more merit than I usually give.
But let's consider what really happens to that 2/3 of the force which I'll call box-moving reactionary forces. Assuming typical construction with an extra-sturdy baffleboard, the force direction on the side panels is in the same direction as the panel, which doesn't deform in that direction. It mostly couples into the mass of the cabinet and into the floor. Let's face it, if there's forces left from that 2/3 they're actually moving the box fore and aft, like the kick of a shotgun LOL.
Quantifying generally the forces that remain to cause problems isn't too difficult. Measuring the movement of the side panels sliding forward and back directions gives a good indication of what remains of the 'box-shaking' 2/3 forces. Movement of the side panels in & out are the result of the pressure changes. And of course if the side panels slide back and forth that radiates no sound. The pressure changes are applied to all sides of the entire cabinet, but the 'box-moving' forces aren't. So measurig the direction of movement of the side panels gives a pretty decent indication of the remaining forces that are not transmitted into the floor.
Thinking about the back panel is more interesting. There the "box-moving" 2/3 forces are in opposite direction to the 1/3 pressure forces on the back panel. So while the air pressure is trying to inflate the box and push out the back panel, the "box-moving" forces are pulling the back panel forward by its edges. So though the forces are in opposite directions and add to bend the panel, the acoustic effects subtract and somewhat cancel.
I guess what I learned is that there may be advantage to deep boxes being able to couple all of the "2/3" force into the floor instead of shaking the box fore and aft.
I imagine for full-range boxes there's all kinds of issues as higher frequencies get transmitted thru the structure of the box. And no doubt my model is overly simplistic and theres' additional beind forces etc. involved.
But I still maintain that though greater forces shake the box fore and aft like the kick of a shotgun, those are more easily handled by the side shapes of standard boxes and coupled into the floor, and the sound radiated out of phase by panels flexing is primarily due to pressure changes inside the sealed sub. The 2/3 box-moving forces as a compression wave thru the side panels need no bracing at all, the panel is wonderfully stiff in that direction. My speakers don't tip forward or back when I play bass, the rectangular side panels don't bend into a rhombus. Nor does the entire speaker move much. But the entire structure does try to expand and contract due to the pressure changes, with that force applied at 90 degrees to the panel. The 'box moving' 2/3 tries to make the box back radaite in-phase with the driver mostly at some resonant frequency of the entire box, not all that bad. But the pressure changes try to make all the sides, top, and bottom radaite in the resonant mode of the panel, radiating out-of-phase with the main output and subtracting sharp narrow bands.
Measuring box motions is starting to sound intriguing...
The reaction to the driver's moving mass is 100%, and that is coupled into the basket and baffle and cabinet. However, the driver's moving mass is small (we usually wish it were smaller) compared to the mass of the basket, magnet, baffleboard, and the rest of the cabinet. Carver seems obsessed with cancelling this. Other manufacturers balance those inertial reactons with via two drivers on opposite sides or even three on the sides of a triangular enclosure. And yes the entire cabinet shakes less, but panel vibration issues remain the same.
But talking about the sound energy and opposite reaction to its propagation, which is probably what I'm guilty of underestimating: Whatever the low conversion efficiency is, like you say: a force we'll arbitrarily quantify as .5 is propagated into the room as useful output, and another .5 causes pressure changes in the cabinet which stuffing and trapped air must absorb or return as spring force. So yes, I'll agree these forces are "out of phase" and their directions add rather than cancel, and the basket/baffle/cabinet strucutre gets the counterforce of both, in twice the quantity 2*.5 = 1. That means that yes, the pressure change force is 1/2 as great as the forces from countering propagated sound, which means the air pressure changes are 1/3 th total force.
Looking at the vector direction of those forces and assuming large wavelengths and smaller box dimensions, when the driver is retracting into the box the basket is pushed in the opposite direction, pulling the box forward. The box is shaking the rear panel by its edges, and as it is moving forward the air pressure on it is increasing, the forces all additive in trying to bend the panel. And these box-moving forces are just as directional as the moving-mass reactions. So maybe the people making subs with drivers on opposite sides of the box or all around the box deserve more merit than I usually give.
But let's consider what really happens to that 2/3 of the force which I'll call box-moving reactionary forces. Assuming typical construction with an extra-sturdy baffleboard, the force direction on the side panels is in the same direction as the panel, which doesn't deform in that direction. It mostly couples into the mass of the cabinet and into the floor. Let's face it, if there's forces left from that 2/3 they're actually moving the box fore and aft, like the kick of a shotgun LOL.
Quantifying generally the forces that remain to cause problems isn't too difficult. Measuring the movement of the side panels sliding forward and back directions gives a good indication of what remains of the 'box-shaking' 2/3 forces. Movement of the side panels in & out are the result of the pressure changes. And of course if the side panels slide back and forth that radiates no sound. The pressure changes are applied to all sides of the entire cabinet, but the 'box-moving' forces aren't. So measurig the direction of movement of the side panels gives a pretty decent indication of the remaining forces that are not transmitted into the floor.
Thinking about the back panel is more interesting. There the "box-moving" 2/3 forces are in opposite direction to the 1/3 pressure forces on the back panel. So while the air pressure is trying to inflate the box and push out the back panel, the "box-moving" forces are pulling the back panel forward by its edges. So though the forces are in opposite directions and add to bend the panel, the acoustic effects subtract and somewhat cancel.
I guess what I learned is that there may be advantage to deep boxes being able to couple all of the "2/3" force into the floor instead of shaking the box fore and aft.
I imagine for full-range boxes there's all kinds of issues as higher frequencies get transmitted thru the structure of the box. And no doubt my model is overly simplistic and theres' additional beind forces etc. involved.
But I still maintain that though greater forces shake the box fore and aft like the kick of a shotgun, those are more easily handled by the side shapes of standard boxes and coupled into the floor, and the sound radiated out of phase by panels flexing is primarily due to pressure changes inside the sealed sub. The 2/3 box-moving forces as a compression wave thru the side panels need no bracing at all, the panel is wonderfully stiff in that direction. My speakers don't tip forward or back when I play bass, the rectangular side panels don't bend into a rhombus. Nor does the entire speaker move much. But the entire structure does try to expand and contract due to the pressure changes, with that force applied at 90 degrees to the panel. The 'box moving' 2/3 tries to make the box back radaite in-phase with the driver mostly at some resonant frequency of the entire box, not all that bad. But the pressure changes try to make all the sides, top, and bottom radaite in the resonant mode of the panel, radiating out-of-phase with the main output and subtracting sharp narrow bands.
Measuring box motions is starting to sound intriguing...
I disagree. You have to look at the energy imparted into the box, and the frequency it is at. Over twice as much energy is imparted into the cabinet mechanically as pnematically.
Active vibration cancellation can be VERY effective, and if utilized can make the pnematically transferred energy the most significant, but that is easily dealt with.
With a subwoofer it is pretty easy to push the boxes resonant signature well above the passband and make them innocuous because they are never excited.
dave
if my 12" woofers arent moving much, but mostly just vibrating, what kind of pressure am I dealing with ?
tho I still got ressonances to deal with
and even at this moderate SPL, bass is still felt my feet, and room pressure is there too
wonder how this can be, when woofers arent visibly moving
tho I still got ressonances to deal with
and even at this moderate SPL, bass is still felt my feet, and room pressure is there too
wonder how this can be, when woofers arent visibly moving
What's the most practical way of implementing a tight bracing for the driver magnet? I'm not too confident of my ability to cut the bracing down to the exact size needed, and don't think there's any point in using squishy material to give some allowance since they won't, IMO, be doing anything.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.