I have been planning to build a speaker with a carbon fiber skin for awhile now, but every mention of using carbon fiber in a speaker cabinet is followed by doomsday type statements, with people saying that it's just really bad.
My gut response is to dismiss those comments, but I decided it is worth researching and testing before I commit to it. Carbon fiber is definitely not common, but I have found 2 examples of speakers using carbon fiber in the cabinets. One is a studio monitor with a carbon fiber skinned front baffle, and the other a high end speaker made from carbon fiber with aluminum honeycomb core.
The plan would be to apply a "skin" of carbon fiber over a cabinet after it is made. The questions I have are: Is it better to skin the outside only, or skin both inside and outside? What is the difference between skinning MDF vs Plywood? And how does a carbon skin compare to a hemp skin? Hemp is promoted as providing "damping" to a composite, and damping sounds good in the context of a speaker cabinet.
To test, I made a bunch of samples of all the combinations I am interested in. They ended up being 6"x 7". I put them in a test jig that "taps" them equally, and I measured the max SPL for a single drop. I took a bunch of measurements and averaged them.
The test jig is pretty simple. It has a board that lifts up against a fixed stop, and gravity swings it down. I set the test pieces on a thin foam strip to prevent rattling against the test jig. This made the results more consistent and repeatable.
Results:
Plywood, bare ACX 3/4": 108.93dB
Plywood with carbon on 1 side: 96.36dB
Plywood with carbon on 2 sides: 94.62dB
Plywood with hemp on 1 side: 105.61dB
Plywood with hemp on 2 sides: 98.5dB
MDF, bare, 3/4": 102.9dB
MDF with carbon on 1 side: 109.8dB (back side impregnated with epoxy b/c otherwise it will warp from moisture)
MDF with carbon on 2 sides: 102.88dB
MDF with hemp on 1 side: 108.86dB
MDF with hemp on 2 sides: 101.38dB
1" extruded polystyrene foam board bare: 110.4dB (taped down so it didn't bounce on jig)
1" extruded polystyrene foam board carbon 2 sides: 106.93dB
I also observed that the resonant frequency of the samples changes when adding the different skins, as one would expect. Tapping bare MDF makes a lower pitch than tapping bare plywood, but adding 1 layer of carbon to the MDF makes the resonant frequency roughly the same as the bare plywood. Adding the second layer increased the frequency again, but only the equivalent of a couple notes on a musical scale.
My takeaway at the moment is that nothing bad happens when you cover plywood or MDF with carbon fiber. The resonant frequency increases due to the panel becoming more rigid; the same thing happens if you put your bracing closer together. It also appears that common pine plywood is a good choice to use with a carbon skin. MDF seems to be more resonant with a skin, possibly because the resin soaks into the MDF and changes it's properties? And the foam didn't do particularly well in this test, which may be due to the fact it is so light weight that it bounces out of the test jig.
I may do more experiments. I was considering using carbon fiber with an endgrain balsa core material. That is denser than the extruded foam, but much less dense than plywood or MDF. It is rigid and light weight, and my guess is it is probably better than the foam.
My gut response is to dismiss those comments, but I decided it is worth researching and testing before I commit to it. Carbon fiber is definitely not common, but I have found 2 examples of speakers using carbon fiber in the cabinets. One is a studio monitor with a carbon fiber skinned front baffle, and the other a high end speaker made from carbon fiber with aluminum honeycomb core.
The plan would be to apply a "skin" of carbon fiber over a cabinet after it is made. The questions I have are: Is it better to skin the outside only, or skin both inside and outside? What is the difference between skinning MDF vs Plywood? And how does a carbon skin compare to a hemp skin? Hemp is promoted as providing "damping" to a composite, and damping sounds good in the context of a speaker cabinet.
To test, I made a bunch of samples of all the combinations I am interested in. They ended up being 6"x 7". I put them in a test jig that "taps" them equally, and I measured the max SPL for a single drop. I took a bunch of measurements and averaged them.
The test jig is pretty simple. It has a board that lifts up against a fixed stop, and gravity swings it down. I set the test pieces on a thin foam strip to prevent rattling against the test jig. This made the results more consistent and repeatable.
Results:
Plywood, bare ACX 3/4": 108.93dB
Plywood with carbon on 1 side: 96.36dB
Plywood with carbon on 2 sides: 94.62dB
Plywood with hemp on 1 side: 105.61dB
Plywood with hemp on 2 sides: 98.5dB
MDF, bare, 3/4": 102.9dB
MDF with carbon on 1 side: 109.8dB (back side impregnated with epoxy b/c otherwise it will warp from moisture)
MDF with carbon on 2 sides: 102.88dB
MDF with hemp on 1 side: 108.86dB
MDF with hemp on 2 sides: 101.38dB
1" extruded polystyrene foam board bare: 110.4dB (taped down so it didn't bounce on jig)
1" extruded polystyrene foam board carbon 2 sides: 106.93dB
I also observed that the resonant frequency of the samples changes when adding the different skins, as one would expect. Tapping bare MDF makes a lower pitch than tapping bare plywood, but adding 1 layer of carbon to the MDF makes the resonant frequency roughly the same as the bare plywood. Adding the second layer increased the frequency again, but only the equivalent of a couple notes on a musical scale.
My takeaway at the moment is that nothing bad happens when you cover plywood or MDF with carbon fiber. The resonant frequency increases due to the panel becoming more rigid; the same thing happens if you put your bracing closer together. It also appears that common pine plywood is a good choice to use with a carbon skin. MDF seems to be more resonant with a skin, possibly because the resin soaks into the MDF and changes it's properties? And the foam didn't do particularly well in this test, which may be due to the fact it is so light weight that it bounces out of the test jig.
I may do more experiments. I was considering using carbon fiber with an endgrain balsa core material. That is denser than the extruded foam, but much less dense than plywood or MDF. It is rigid and light weight, and my guess is it is probably better than the foam.
Interesting. You could try kevlar, which has higher damping than carbon - either as the skin alone, or laminated with a layer of carbon too.
I didn't try kevlar for aesthetic reasons. I don't want yellow speakers! I do have some Kevlar laying around, though...
As far as damping, I believe the hemp has the highest damping of laminating fabrics. So the damping of kevlar would be less than the hemp.
I don't think either hemp or kevlar would effectively "damp" a panel. The hemp increased the resonant frequency, it did more to make the panel stiffer than it contributed any damping.
Thanks and congratulations on actual testing 🙂
Never tried carbon fiber (yet) but it´s a tough modern material, so why not?
back in the day covered lots of cabinets with fiberglass + resin (usually polyester because of cost) and results were excellent,only drawback is that it increased enclosure weight considerably, but that should be less with thinner stronger carbon fiber and concurrent less amount of Resins.
Keep us updated.
You can add pigment to epoxy to colour it, which can look quite nice. Or just paint it after, of course.
I suppose kevlar would be somewhere between hemp and carbon for both damping and stiffness then.
I would go for choosing one material for stiffness and another to provide damping, stiff materials are usually very resonant. Stiffness and mass can reduce the response to stimulus, and their ratio (plus geometry) determines resonant frequencies, but only a lossy material can improve damping to quell those resonances. Elastomers are often used for damping as even a thin layer can provide a lot of effect especially in a laminated structure. For instance the three thin rubber layers in a cricket bat handle absorb a lot of shock/vibration from the impact that would otherwise transfer to the players hands.
JBL LSR32 have carbon fiber baffles, pretty highly regarded. However the later 6332 identical speaker does not.
The pro version of the IMF RSPM IV came with a formica skin which apparently made a drastic improvement.
The trick with MDF is to apply primer and sand to water break free finish. Then apply some paint that will sit well behind the carbon fibre. Sand this to water break free level too. This now makes a base for applying cloth. The stitched CF cloth is not for this job for all the reasons already mentioned. Use woven as all it does is add weight and looks
When you apply cloth to both sides, you create a sandwich where less stretchy material has to bend around a thickness if asked to and this creates the new stiffness. This makes for a stronger enclosure but I have never given thot to resonant frequency and give respect to your original initiative. Use thin walls and stitched biax glass cloth on both sides for strength and an outer cosmetic woven glass cloth for the looks that you want. Then apply some car audio dampening for inside. MDF is good for this if machined but ply is developable
When you apply cloth to both sides, you create a sandwich where less stretchy material has to bend around a thickness if asked to and this creates the new stiffness. This makes for a stronger enclosure but I have never given thot to resonant frequency and give respect to your original initiative. Use thin walls and stitched biax glass cloth on both sides for strength and an outer cosmetic woven glass cloth for the looks that you want. Then apply some car audio dampening for inside. MDF is good for this if machined but ply is developable
I tried a laminate of ~1/4" cork, between two ~1/4" sheets of plywood. A 24" X 24" plate still knocks like a door in a corner...
I'll guess a layup of alternating stiff and soft, absorptive materials could make a pretty non-resonant panel. They have rubber sheet for hooves to rest upon at the local farm supply that would count as "absorptive", but with the thicknesses they have at 1/4, 1/2, and 3/4" you could only realistically do one layer constrained by two stiff ones.
Butyl rubber has been around for a while; one would think somebody sometime would have tried it in a 3/4" board layup like w-r-w-r-w-r-w but perhaps building a sheet of something like that would be cost prohibitive. What would you use for the adhesive? How would you get it to dry? I used 3M countertop laminate spray for the cork to plywood bond...
I'll guess a layup of alternating stiff and soft, absorptive materials could make a pretty non-resonant panel. They have rubber sheet for hooves to rest upon at the local farm supply that would count as "absorptive", but with the thicknesses they have at 1/4, 1/2, and 3/4" you could only realistically do one layer constrained by two stiff ones.
Butyl rubber has been around for a while; one would think somebody sometime would have tried it in a 3/4" board layup like w-r-w-r-w-r-w but perhaps building a sheet of something like that would be cost prohibitive. What would you use for the adhesive? How would you get it to dry? I used 3M countertop laminate spray for the cork to plywood bond...
Placing the dampening material inside the sandwich does this, adds more thickness around which a non-stretchy cloth will bend around to bend, so makes it stiffer. The stiffness comes from skin adhesion and material gap between the two skins. That's why I suggested dynamat over the sandwich and only on the inside face. Cork is too light for dampening like that. Remember that the best rod grips are cork to transmit bite from rod to hand
Good glue for cork is any solvent based clear non-expanding PU, epoxy with a light mix of hard wood sanding waste is better
Another suggestion I have is to try doping over your laminate on the inside surface. Apply some nylon cloth about 1mm thick and slather on acrylic paint then further damp with dynamat. I make glass consoles and pods for all sorts of things from marine to ornamental
Good glue for cork is any solvent based clear non-expanding PU, epoxy with a light mix of hard wood sanding waste is better
Another suggestion I have is to try doping over your laminate on the inside surface. Apply some nylon cloth about 1mm thick and slather on acrylic paint then further damp with dynamat. I make glass consoles and pods for all sorts of things from marine to ornamental
I paint the insides of my speakers with this: EVERBUILD® BLACK JACK® 908 D.P.M. | Bitumen & Roofing Products
Originally stumbled across it looking for a way to attach acoustic foam panels without running into chemical problems (this is water-based and therefore practically inert once dry) which it does beautifully but it also sets into a seamless heavy rubber coating which has got help dampening although I never actually tested it like you did. The ultimate thickness depends on how many coats you can be bothered to apply.
Originally stumbled across it looking for a way to attach acoustic foam panels without running into chemical problems (this is water-based and therefore practically inert once dry) which it does beautifully but it also sets into a seamless heavy rubber coating which has got help dampening although I never actually tested it like you did. The ultimate thickness depends on how many coats you can be bothered to apply.
The one thing that always fascinated me about the use of synthetic fiber either on cabinets or in speaker cones is the application of pre-stress. If the resistance to deflection is highly asymmetrical then you might get the idea that second order noises at fundamental would dominate. If the tension were dialed up to say half the limit of the material then maybe apparent strength would be much higher than an un-laminted sheet. You could stuff in braces that are just a little bit too long to do that for any cabinet but in the case of natural materals especially like MDF it would eventually deform permanently to relax the original stress, losing the plan.
Best set of La Scala Industrials, well make that the best La Scalas period I have ever owned were covered in fiberglass. It seemed to work well to dampen cabinet resonance though that may have been a function of thicker sidewalls then any other thing. I use Baltic Birch for my speaker builds and like the tone and don't have to cover them in anything to sound good. Of course if this is something you just want to do for the heck of it that is a good enough reason to try.
I'm planning to use some 4HS fabric I have laying around, and it doesn't have many gaps in the weave. But I completely agree on having a black layer under the fabric, just in case! I was thinking add dye to the resin when sealing the wood for the black under coat
I was planning to seal the cabinet with resin before I applying the fabric. I'd seal it, then sand and fill as required, and then apply the fabric. I'll apply fabric to top and sides of the cabinet with baffles removed, and I plan to laminate the baffles separately. I'll then bond the baffles to the cabinet, seal any gaps with epoxy and carbon paste. Then apply a clear resin to the assembled speaker, and finish it out. Probably finish with a urethane clear for good measure.
The skin is all I'm really going for. I'll put some carbon inside just to balance it, but I won't put all that much effort into the interior.
I think you're describing "constrained layer damping"? Two materials separated by a viscous material that limits transmission of vibrations. It is a thing. However, the effect will depend on the materials chosen, and it is well beyond what I'm trying to do! You'd have to find products made specifically for the purpose, and construction may be a bit challenging.
The premise of my post was to see if a carbon fiber skin will ruin a speaker or have some negative effect. It isn't much of a test, but it is enough to make me think that it'll be just fine!
They make fantastic canoes and such using just western red cedar veneer & epoxy (West System if you want to search it out)
dave
dave
I have thought about a laminate using balsa core material. It would be cool, but in my test the plywood covered with carbon seemed the best. And it's tougher than balsa, which is important since I'm hoping to use a single layer of carbon. (Using balsa would offer welcome weight savings. The current prototype uses 1.5 sheets of chip board. I tweaked my back wrestling with it this weekend!)
The damping. I don't think mass loaded material works in the center of the laminate. The scenario where you put a damping material in the middle is constrained layer damping, where you use a viscous material to de-couple the two outer layers. The mass loaded material would be placed on the surface for it to work.
I got my start with fiberglass building cedar strip canoes! I have built several. They were all strip built, so I never tried veneer or plywood. I would use a technique like that if I was building a complex shape.
But right now I am liking the idea of the speakers being essentially pillars, a bit monolithic. I posted a couple photos of the most current prototype on a different thread. Right now it's 18"x18"x48". I think it is a bit too wide. 16" wide would probably look about right.
https://www.diyaudio.com/community/...-area-on-low-level-output.391733/post-7167420