It's a really mixed bag.
Some do everything in house, incl electronics, speaker design and even complete box manufacturer.
If we just talk about wood, a lot of them get semi-assembled and finished cabinets.
Polpar iss one of the most damped wood, very ligth... But indeed not stiff enough. Poplar would be good in a sandwich internal layer with more stiff wood.
Oak is not good because too stiff and too dense (lack damping).
Anyway, the number of layers is helping in a plywood cause mass spring mass effect. The more the better. 13 layers plywood is ok as a mix of stifness damping and mass (weigth) combos.
I picked up a few sheets of 3/4” birch ply that my supplier just got in, its made in Taiwan is about all I know about it besides it looks/feels pretty decent for the money ($48/sheet) i‘ll do some forensics on it tomorrow when I break it down.
Going to use it to build some proto-types…….its nowhere near baltic birch quality but we‘ll see how it goes.
Going to use it to build some proto-types…….its nowhere near baltic birch quality but we‘ll see how it goes.
Well upon further inspection its 13 ply and seems well glued with minimal voids, BUT……it‘s a very very thin outer layer of birch that is paper backed! First glance at the supplier looked like 1/32” or so of veneer…….nope. Good enough for prototypes at less than half the cost of marine grade fir but not end game.
13 layers should make it stiff and well damped. It will be better to know the wood(s) variety. But imo if above 700kg/m3 you are good not only for the proto.
HumbleHomeMadeHifi use that thickness for its last 105 liters bass cabinet with a 12" . If you use some bracing inside with CLD technique it should be more than enough I surmise.
HumbleHomeMadeHifi use that thickness for its last 105 liters bass cabinet with a 12" . If you use some bracing inside with CLD technique it should be more than enough I surmise.
Upon dissection the white wood is definately poplar (can see the green heart wood) but that brown stuff is a hard heavy wood almost like teak but surely not? I’m not too familiar with Vietnamese woods, but they did a pretty good job manufacturing it.
Attachments
Perhaps Mahogany red wood that has the same spec than poplar...Quite comon when made in China for instance.
If you feel it is too ligth when knocking it. Doubling with 1 cm OSB (very stiff, heavy and damped) behind or 0.5 mm HDF on the front baffle is surely ok. Also 3/4 hdf for the front baffle standalone too. Use CLD bracing technic for bracing to avoid coupling vibrations between the cabinets sides.
If you feel it is too ligth when knocking it. Doubling with 1 cm OSB (very stiff, heavy and damped) behind or 0.5 mm HDF on the front baffle is surely ok. Also 3/4 hdf for the front baffle standalone too. Use CLD bracing technic for bracing to avoid coupling vibrations between the cabinets sides.
Regarding excessive bracing: Yes it might be possible, in theory, to have so much bracing in a cabinet that other problems are created, such as the internal air flow is impeded. Your post, however, creates the impression that there is actually a risk of this happening. I strongly doubt that excessive bracing something that happens very often, if ever.
.... And it is a complicated subject, because there are many successful, low-signature cabinets which are not particularly stiff, but instead use structural damping methods. The fact that there are multiple contradicting techniques which can lead to success is an indication that the cabinet's structural effect on listener preference is complicated and not very well understood.
j.
Many successful, low-signature cabinets have been built which are extremely rigid. A very stiff cabinet has higher structural resonance frequencies, they are higher Q resonances, and there are fewer of them. High Q structural resonances are less likely to be energized by program material, and listening evaluations have shown that low Q resonances are more audible than high Q ones. There is less energy available at high frequencies than at low frequencies in program material. This is why a high frequency resonance which is measurable with white noise or sine sweeps is often insignificant, while a low frequency resonance detected with the same methods is often audible.
.... And it is a complicated subject, because there are many successful, low-signature cabinets which are not particularly stiff, but instead use structural damping methods. The fact that there are multiple contradicting techniques which can lead to success is an indication that the cabinet's structural effect on listener preference is complicated and not very well understood.
j.
I think the listening preference is the last item that matters for audiophile companies. The subject for them is just about their price market niche / cost ratio then is talked in the marketing side. Rigid: see Rockporttechnologies, flexible: see Harbeth.
They all know you listen first your loudspeaker at home with eyes and brain.
Professional market is something else and customers are looking for reliability to cost to use efficienty.
I am not surprised poplar can be used : it is ligtht, flexibility helps when a box falls or is shaked by transport... Vuitton malles are made with poplar from day one for those reasons... It is resiliant.
If you need rigidity and space matters then you have to increase the layers number and reduce their thickness.
We also know density and weigth help when frequencies go lower. A 5 cm concrete wall is isolating less at low frequency where the wavelength is big than a 20 cm wall. Here a Panzerholtz ply is good VS Ply but at what cost (monney and weigth increase greatly)
What we ideally look for is external sides that do not sing and internal volume that damp enough not to interact too much on the back of the cone. And not too much to avoid too muxh stuffed sound. Plus internal damping to avoid external cabinet sides to sing. Sandwich and mass spring mass effect to loose energy is a well known technic and CLD construction as sandwich cabinet reflect that.
When someone here is saying a two layers mdf glued together is better than a multiple ply, it is certainly wrong. He dosen't understand it is just a 2 ply layers made from low density particles. Ok it is better than a singlle layer.
But when you have multiple layers you increase the energy loss by increasing the mass spring mass effect as you increase the rigidity at the same time. Ideal should be to have a ply with different layers thickness and material I surmise. Here by the way, OSB, according how they are made and wood choice is an ok choice. But a pain to manufacture for cabinet makers... here comes the Mdf... so easy and cheap ! HDF > 750 kg/m3 looks good though : heavy and rigid enough, damping better than MDF. Used for front plate for instance.
Looks like that is what happen inside that is more complicated to solve cause the resonances and multiple frequency lengths as well.
The second box being the room that leaks with the listener inside. Then you realise it... the cabinet is becomming less a concern.
They all know you listen first your loudspeaker at home with eyes and brain.
Professional market is something else and customers are looking for reliability to cost to use efficienty.
I am not surprised poplar can be used : it is ligtht, flexibility helps when a box falls or is shaked by transport... Vuitton malles are made with poplar from day one for those reasons... It is resiliant.
If you need rigidity and space matters then you have to increase the layers number and reduce their thickness.
We also know density and weigth help when frequencies go lower. A 5 cm concrete wall is isolating less at low frequency where the wavelength is big than a 20 cm wall. Here a Panzerholtz ply is good VS Ply but at what cost (monney and weigth increase greatly)
What we ideally look for is external sides that do not sing and internal volume that damp enough not to interact too much on the back of the cone. And not too much to avoid too muxh stuffed sound. Plus internal damping to avoid external cabinet sides to sing. Sandwich and mass spring mass effect to loose energy is a well known technic and CLD construction as sandwich cabinet reflect that.
When someone here is saying a two layers mdf glued together is better than a multiple ply, it is certainly wrong. He dosen't understand it is just a 2 ply layers made from low density particles. Ok it is better than a singlle layer.
But when you have multiple layers you increase the energy loss by increasing the mass spring mass effect as you increase the rigidity at the same time. Ideal should be to have a ply with different layers thickness and material I surmise. Here by the way, OSB, according how they are made and wood choice is an ok choice. But a pain to manufacture for cabinet makers... here comes the Mdf... so easy and cheap ! HDF > 750 kg/m3 looks good though : heavy and rigid enough, damping better than MDF. Used for front plate for instance.
Looks like that is what happen inside that is more complicated to solve cause the resonances and multiple frequency lengths as well.
The second box being the room that leaks with the listener inside. Then you realise it... the cabinet is becomming less a concern.
Last edited:
I'm not sure I agree with that. If two layers of MDF are glued together with the right kind of glue, using the right kind of technique, the result is a constrained layer damped panel (CLD). In a high-damping cabinet with minimal bracing, this can be a significant step up from an equivalent thickness of plywood.
The glue bond in plywood is so strong that I doubt there is any meaningful damping from the bond lines between layers. I think the internal damping in plywood comes for the wood itself.
Well I agree with this. Cabinet radiation does not need to be zero, it just needs to be low enough. There is a lot of ways to get there.
If you have only two layers there is no shearing effect but "just mass spring mass effect. Shearing effect is certainly small but existing with ply, plus multiple in "serie mass spring mass" effect is what is what imo makes it superior. This is what you call "comming from the wood itself" and it onlyy works because there is a layer of thin glue that has a different density than the wood layers themselves (mass spring mass effect).
A two layers Mdf will be less efficient with two same thickness layer because it will be centred on a narrow frequencies window and anyway willl be less stiff than ply where each layer has wood sense at 90° - so it is not the wood only but how the ply is made-. But of course better (already said) than a standalone that has same whole thickness. But only cause the mass spring effect. It is not sure it is more rigid at iso thickness because the particles have no structural sense and is just saw dust mixed with glue and made compact. You noticed that byy doubling a thin ply with a sheet of thin mdf on your site. And it works fine because the two layers have not the same internal damping.
Note that in CLD the glue layer must be thin, which make it not easy for use because we have not heavy press and Anyway it works better with 3 layers, the inner layer just coupled by glue to the two others that have structural contacts with the rest of the cabinet. Here an ultra damped wood like poplar (low density wood) would work the best. I surmise for this shearing layer between two more dense wood, a basic mdf should work fine.
It will be clever to know enough to make the sonic external resonance of the cabinet in the area of room mode where there is a null and not a peak. Like for instance 300 hz or around 150 hz where there are often spl null because the layout of the speakers vis à vis of the front wall. Maybe... I do not know. But notice litterature is saying we shoukd make it higher in frequency where it is said armless because less strong (wave length?)
I often wonder how those vibrations are measured by Stereophile. Is is just with a sweep or tone bursts ? The second looks like more efficient maybe to evaluate how the external sides of the cabinet are singing ?
A two layers Mdf will be less efficient with two same thickness layer because it will be centred on a narrow frequencies window and anyway willl be less stiff than ply where each layer has wood sense at 90° - so it is not the wood only but how the ply is made-. But of course better (already said) than a standalone that has same whole thickness. But only cause the mass spring effect. It is not sure it is more rigid at iso thickness because the particles have no structural sense and is just saw dust mixed with glue and made compact. You noticed that byy doubling a thin ply with a sheet of thin mdf on your site. And it works fine because the two layers have not the same internal damping.
Note that in CLD the glue layer must be thin, which make it not easy for use because we have not heavy press and Anyway it works better with 3 layers, the inner layer just coupled by glue to the two others that have structural contacts with the rest of the cabinet. Here an ultra damped wood like poplar (low density wood) would work the best. I surmise for this shearing layer between two more dense wood, a basic mdf should work fine.
It will be clever to know enough to make the sonic external resonance of the cabinet in the area of room mode where there is a null and not a peak. Like for instance 300 hz or around 150 hz where there are often spl null because the layout of the speakers vis à vis of the front wall. Maybe... I do not know. But notice litterature is saying we shoukd make it higher in frequency where it is said armless because less strong (wave length?)
I often wonder how those vibrations are measured by Stereophile. Is is just with a sweep or tone bursts ? The second looks like more efficient maybe to evaluate how the external sides of the cabinet are singing ?
The stiffness in plywood is mostly because of the different grain orientation with each layer.
So one layer goes into x direction, the other layer goes into y direction.
MDF has no grain orientation and is basically just saw dust mixed with a binder to be formed into some kind of panel.
Because of the lack of any grain orientation, I have a hard time seeing how MDF could be more stiff even when this has been layered with glue?
So one layer goes into x direction, the other layer goes into y direction.
MDF has no grain orientation and is basically just saw dust mixed with a binder to be formed into some kind of panel.
Because of the lack of any grain orientation, I have a hard time seeing how MDF could be more stiff even when this has been layered with glue?
MDF becomes only denser with more cubic saw dust at iso volume, it basicly is allowed by increasing the force of the press when they make the pannels
Garnica talked in post one is an excellent brand for making HDF and tinted in the mass, that is what I think T Gravsen uses for some black front plate.
A cheap mdf is 500 to 600 kg/m3 density and is not so ok for big pannels... certainly is for lower mid or treble splited cabinet though or mini 2 ways.
HDF becomes ok with more than 700/750 kg/m3.
For diy, OSB class III inside and basic 0.5 mm MDF outside for clean looking finition would be my small pockett choice for good enough.
Garnica talked in post one is an excellent brand for making HDF and tinted in the mass, that is what I think T Gravsen uses for some black front plate.
A cheap mdf is 500 to 600 kg/m3 density and is not so ok for big pannels... certainly is for lower mid or treble splited cabinet though or mini 2 ways.
HDF becomes ok with more than 700/750 kg/m3.
For diy, OSB class III inside and basic 0.5 mm MDF outside for clean looking finition would be my small pockett choice for good enough.
With all due respect, I don't want to be argumentative... but some of the things you are saying are not true, assuming I understand you correctly.
Two stiff layers (plywood, mdf, etc) separated by a viscous layer of semi-soft glue will have significantly more damping than a solid layer of the stiff material. The glue layer will undergo far more sheering deflection (in the plane of the material) then the stiff layers will undergo. It is the deflection of the visco-elastic layer that dissapates energy in a CLD panel.
Actually it must be much thicker than a typical adhesive bond line. @augerpro has recommendations for this, and if I recall the recommendation is for a final viscoelastic adhesive thickness of between 1/32 and 1/16 inch (about 1 mm). Most adhesive bond lines would be much thinner than that.
Lets talk about straight grained wood timber of the sort used in load bearing applications (such as pine, fir, oak, birch, beech, mahogany, and many other hardwoods). The timber is 3x to 5x stiffer in the direction of the grain versus across the grain. Plywood has alternating plies. The intent of plywood is make a sheet product which has the same stiffness and strength in any direction, and this is typically in the range of 0.707x the stiffness / strength of the timber in the strong direction. The layering does not add stiffness, it reduces it, but it makes it the same in every direction.
Again, I am not trying to be argumentative, but it is rather important to have good information out there for future reference.
j.
Do I said than two layers have equall damping than a single one? I talked about stiffness of two VS one at iso whole size. Here I talk of bonding not damping.
Damping : again the increasing damping effect occurs because of several densities that break the WL - not only brake and reduce it at a frequency according the thickness of a standalone thickness and material combo. Marco Gea said it all about Young Modulus about a material. plain oak is for instance less wanted than a ligther wood because a little too dense for our hobby to make plain panels
Your two layers will be stiffer only if they are structurally both tied with the top and bottom cabinet sides as also the adjacent sides. Understand : the inner layer in the cabinet is not floating it will be stiffer. But what you say is two mdf glued are stifer than an iso total mdf sheet thickness if I understand you. I.e. : two layers of 1 cm glues should be stiffer than one of 2 cm. I think it is wrong when talking about mdf. If two layers of 1 cm is stiffer than a standalone 1 cm one, then rigth there is no discuss !
Then my understandinng is there is no shearing effect if it is not floating. Just mass sringg mass effect. The glue itself is not shearing if I am rigth but permit the shearing of the other layer that is floating. If you want true cld in your cabinet if having two structural layers with the thickness sides of one of it in the not glued (but just the plain area sides). Or use a third internal layet that floats !
I dunno what is a bondline... my english. I noticed after used those elastic glue (Sykal) that it is very hard to apply then press with uniformity. And without pro press I dunno how one can experiment with diferrent thickness of glue arounf the mm ?! But if you are saying it must be thicker than the glue layer of a ply, I surmise you are rigth. But notice those are not the same glue, the one used for CLD must be visco elastic (understand must have a resilience effect after displacement)
I am ok with the argumentation than ply is stiffier because it is made by crossing the layers, it is indeed stiffier in every direction. Which is wanted when having a big cabinet surface : it doesnt only bond in the bigger length if the pannel is big enough. Also and you are saying it : it will be stiffier only in the grain of wood fiber direction (and also make it less damped than the ply so more resonant at some frequencies). It is about precise information imo.
Damping : again the increasing damping effect occurs because of several densities that break the WL - not only brake and reduce it at a frequency according the thickness of a standalone thickness and material combo. Marco Gea said it all about Young Modulus about a material. plain oak is for instance less wanted than a ligther wood because a little too dense for our hobby to make plain panels
Your two layers will be stiffer only if they are structurally both tied with the top and bottom cabinet sides as also the adjacent sides. Understand : the inner layer in the cabinet is not floating it will be stiffer. But what you say is two mdf glued are stifer than an iso total mdf sheet thickness if I understand you. I.e. : two layers of 1 cm glues should be stiffer than one of 2 cm. I think it is wrong when talking about mdf. If two layers of 1 cm is stiffer than a standalone 1 cm one, then rigth there is no discuss !
Then my understandinng is there is no shearing effect if it is not floating. Just mass sringg mass effect. The glue itself is not shearing if I am rigth but permit the shearing of the other layer that is floating. If you want true cld in your cabinet if having two structural layers with the thickness sides of one of it in the not glued (but just the plain area sides). Or use a third internal layet that floats !
I dunno what is a bondline... my english. I noticed after used those elastic glue (Sykal) that it is very hard to apply then press with uniformity. And without pro press I dunno how one can experiment with diferrent thickness of glue arounf the mm ?! But if you are saying it must be thicker than the glue layer of a ply, I surmise you are rigth. But notice those are not the same glue, the one used for CLD must be visco elastic (understand must have a resilience effect after displacement)
I am ok with the argumentation than ply is stiffier because it is made by crossing the layers, it is indeed stiffier in every direction. Which is wanted when having a big cabinet surface : it doesnt only bond in the bigger length if the pannel is big enough. Also and you are saying it : it will be stiffier only in the grain of wood fiber direction (and also make it less damped than the ply so more resonant at some frequencies). It is about precise information imo.
Last edited:
If you use the right trowel you can get a uniform adhesive thickness.
FYI I used 1/32" thickness because that was about what I remember Geddes recommending.
FYI I used 1/32" thickness because that was about what I remember Geddes recommending.
I may be wrong but still think than a 13 layers ply of 2 cm is stiffier but also more damped than two 1 cm MDF glued together whatever the glue.
And that there is not shearing effect with two mdf panels glued together if the two panels are both structurally glued to the others sides of the cabinet. My understanding is for a shearing effect talked in CLD technic : one panel must freely vibrate without transmit it to the others adjacent panels but with the only one it is glued on (one side only and the biggest surface one). It is for a CLD with two pannels : so the second pannel that vibrate is not touching the cabinet itself.
I may not understanding what CLD is but imho the guy than make it rigth for pannel doubling is @mbrennwa with an internal aluminium floating panel that is not too much thick and that dissipate heat easily.
Bracing is another subject and if my understanding is stilll correct the shearing effet (CLD) will be optimised if there is three layers where the inside layer is floating : just attached to the two other bottom and top braces that are itselves tied with the opposite panels sides. Most of the energy will ne dissipated into heat in that internal free layers (just in contact with visco elastic glue).
And that there is not shearing effect with two mdf panels glued together if the two panels are both structurally glued to the others sides of the cabinet. My understanding is for a shearing effect talked in CLD technic : one panel must freely vibrate without transmit it to the others adjacent panels but with the only one it is glued on (one side only and the biggest surface one). It is for a CLD with two pannels : so the second pannel that vibrate is not touching the cabinet itself.
I may not understanding what CLD is but imho the guy than make it rigth for pannel doubling is @mbrennwa with an internal aluminium floating panel that is not too much thick and that dissipate heat easily.
Bracing is another subject and if my understanding is stilll correct the shearing effet (CLD) will be optimised if there is three layers where the inside layer is floating : just attached to the two other bottom and top braces that are itselves tied with the opposite panels sides. Most of the energy will ne dissipated into heat in that internal free layers (just in contact with visco elastic glue).
Last edited:
I wasn't thinking about heat dissipation when I did this with my OSMC speakers. The idea was that sound propagates at different speeds in wood and metal. At their given panel dimensions, the two materials would therefore resonate at different frequencies. If one resonates, the other would tend to keep still, making the whole thing less sensitive to panel resonances.
That said, THE best way to deal with the problem is bracing. Bracing, bracing, bracing!