Effect of cone material on low level resolution?

[azrix]

Hello all. This isn't the easiest topic to search for but I have searched and I haven't found any threads specifically dealing with this topic, so please point me in the right direction if it's been covered before...

I've been working on my own three way design (my first project) with a general idea about how it's going to be and trying to find info to help with choosing which drivers to go with, with low distortion and ability to convey low level detail a top priority in deciding. I used to think I was going to go with an Accuton/Thiel tweeter and mid driver but after reading the thread "Best midrange for intelligibility of voice" and what Dan Wiggins said (pgs 3-6) about excursion exciting the resonances of stiff cones, I'm rethinking things. I used to think low enough or steep enough lowpass filters would make stiff cones work, like how Joseph Audio does it with their 120db crossover (an extreme case). I see now that isn't the case.

My listening experience is limited, but I've read many, many times in forums and reviews how polypropylene [polymer] cones just don't have as much low level resolution that stiff cones do (speaking mostly of bass/mid woofers). I'm just wondering if this is true all the time or just part of the time and is dependant on motor or cabinet/box design, driver construction quality, etc. I'm speaking of the better and mostly more expensive drivers: for stiff cones I'm thinking of Accuton, Eton, and Seas Excel metal drivers, and for poly cones I'm thinking of Morel, Skaaning/Audio Technology, the Adire Extremis, and Dynaudio (which I'd like to include in this discussion). There are others I know, so feel free to talk about them, but I was mostly looking at these. I'd also ruled out paper cone drivers for the midrange because of other things I've read about them (right or wrong). I know the pro's of poly drivers and have thought the lack of detail was it's main con, but now I'm not so sure. I think I can live with the cons of stiff/metal cones if they have more detail, but I don't really know that for sure and was hoping for some answers.

So, the question is: Given high quality drivers, are stiff (metal, ceramic, carbon) cone drivers better able to resolve detail than poly cones? Why or why not?

Thanks.

Ben.

 

[David Gatti]

Good post azrix.
An interesting comparison is made by Mark K between the Seas magnesium (w15) & poly (M15) cones.
http://206.13.113.199/ncdiyaudio/mark/Midrange test group/midrange_test_data.htm

The findings are interesting.
Basically, below 1kHz, the metal cone has superior linear distortion, but from 1kHz up, the M15 was superior.

As for non linear distortion, the W15 did better at low frequencies (possibly due to different motor), and higher up it was slightly better, but nothing to get excited about.

I think poly/paper cones get a bad rap because there are many cheap designs out there using these economic & easy to handle materials. That doesn't mean that they should all be labelled as poor.

Also most good polypropylene & paper cones are hybrids loaded with another material so they shouldn't all be grouped as one .

David

 

[David Gatti]

It's interesting to note that Seas's latest Excel drivers are actually paper! Makes one wonder.

http://www.seas.no/Excel line up.htm

 

[Salas]

From what I have gathered by experience and reading, there is a contadiction in physics of what we ask for a cone to optimally do.
Being the lightest and strongest possible as long as being well damped. In real terms we cant have our cake and eat it too.
The stiffer and lighter we go the more resolution and spl per volt we get. Alas as the frequency goes higher, strong resonances strike bcs of insufficient damping of rigid materials.
There have been many composite materials and techniques devised to battle the problem. One good composite was Audax's Aerogel in its maturity. But Audax stopped open trade. Went totally OEM under Harman control.
The best performance that I have ever seen was in a couple of 8 inch experimental midwoofers by Audax that never got into production. A friend got em for evaluation and still has em. They are made from woven ceramic fibre! They have sugar-beige cone color. These really have top resolution and fantastic breakup control both in measuring and auditioning. dont know why they never made it into production. Maybe cost and/or fragility.

 

[Noksukau]

Yeah, it's a shame that many manufacturers are going OEM only now.

As David said, a lot of paper and polypropylene drivers are frowned upon because they are associated with cheaper units but the number of high quality drivers using both materials tells a different tale. (edit: either material, haven't seen anyone use both yet!)

The Vifa wood pulp cones (exactly how this differs from paper I'm not sure ) are supposed to have excellent resolution, in a large part aided by their low moving mass. I've heard a production variant in a Vifa partner loudspeaker and was truly hooked by the natural sound and effortless detail on offer. But then I've never heard them against Accuton or (old) Seas to make a drect comparison. It always seems that the more specialist cone materials need a lot of crossover work to get the most out of them and I dislike complex solutions. Fostex also have a strong pedigree with paper technologies and most of their units are designed for imaging, voice and detail fanatics, not bass-heads. I know you were mainly asking about polypropylene units vs. more 'engineered' materials, not paper, but I wouldn't discount either.

I don't know which is the best, I don't know that anyone does. Heterogenous materials will of course have fewer problems with resonances and break-up and if these aren't masked in a metal/ceramic cone then despite an apparent advantage they won't be as clear. I still think a ceramic cone would be more detailed but I'd prefer to listen to an expertly mixed paper allied to a good motor assembly. That is just personal taste though, which is more important than engineering theory. When I first started out to become a materials engineer I was desperate to work in the audio community and develop my own revolutionary cone material. The more you look at natural materials though the more you realise how well balanced their acoustic properties are. Nature has done a pretty good job already, but I'd still like to see more advanced materials in newer configurations. For example, using multiple alloys to suppress each other's break-up modes (like wood/paper), and arrange them in a corrugated film for extra mechanical stiffness. A single sheet of anything, formed into a cone looks ridiculously flexible from any mechanical standpoint, analogous to building a bridge by spanning a gap with a single beam. At least we're bracing cabinets these days!

Still, getting money to research anything is hard enough for companies and harder still in a niche like high-end audio when, for most people, the top-end solutions already in existence are essentially 'perfect'. For the rest of mankind, a Sony mini-system is audio heaven, because crazy fools that they are, they are more interested in dancing to the music rather than worrying about whether that last cymbal crash highlighted the ugly break-up of their shiny logo-adorned aluminium woofers. Not that I'm blaming any one of us for not loving music, it's our raison d'etre, despite getting lost in the minutiae of its reproduction. After all, without thinkers and questioners in life there wouldn't even be audio reproduction to develop, let alone get obsessive about! I suppose we'd all be too busy going to concerts, trying to learn a fourth instrument, and wooing our loves with cosy recitals and serenades. Thanks to Thomas Edison though we can nip home, slip a Barry White CD into our new-age phonographs AND dim the electric lights! So much easier. Plus, we can content ourselves that mastering the design of loudspeakers is simply learning to craft and play our very own musical instruments. (Or is that blowing one's own trumpet? )

That's what happens when you can't answer someone's question and don't have a point. A big belgian waffle with extra syrup, mmmm tasty!

Ben.

 

[azrix]

Good post azrix.
An interesting comparison is made by Mark K between the Seas magnesium (w15) & poly (M15) cones.
http://206.13.113.199/ncdiyaudio/mark/Midrange test group/midrange_test_data.htm

The findings are interesting.
Basically, below 1kHz, the metal cone has superior linear distortion, but from 1kHz up, the M15 was superior.

As for non linear distortion, the W15 did better at low frequencies (possibly due to different motor), and higher up it was slightly better, but nothing to get excited about.

Yes, very interesting. This quote at the bottom of that page is very telling: "While the W15 nonlinear plot looks great above, increasing the drive level just one more dB caused the driver to start audibly distorting with some rub/buzz type noise."

They mention the same effect with the W18 at 80hz. Looks nasty.

It's interesting to note that Seas's latest Excel drivers are actually paper! Makes one wonder.

http://www.seas.no/Excel line up.htm

Yeah. I've noticed that too. Seas doesn't yet have distortion graphs for these drivers like they do with the rest of the Excel line though.

Ben.

 

[ScottG]

http://www.fostex.com/support/pdf/fostex/nf1/nf1_tech_white_paper.pdf

starting at page 6..

note however that many poly-based loudspeakers are not "pure" poly designs that could contain:
1. structural differences (honycomb, woven, etc.)
2. coatings
3. core differences (sometimes graphite)
4. suspended substrates (sometimes graphite, or other carbon forms, aluminum, or magnesium)

 

[mike.e]

Somewhere on linkwitzlab site there is serious measurements of mid woofers

 

[David Gatti]

Here it is. http://www.woodartistry.com/linkwitzlab/mid_dist.htm
Personally I would never use an 8" polycone (like the Vifa P21) above about 1kHz simply because the heavier cones usually to exhibit stored energy at the top end. The midrange is best suited to a small (4" to 6") cone with low moving mass.

 

[azrix]

Mike.e, are you saying that non-linear or linear distortions as measured by Linkwitz are a good gauge for resolution? I can see the non-linear and other THD tests as being useful for gauging resolution given masking effects, but I'm not sure I see how linear distortions would affect it. I've been thinking resolution is somewhat more complex than that. Is there some single measurement that you think corelates to low level resolution, or at least the perception of that resolution?

Originally posted by ScottG http://www.fostex.com/support/pdf/f...white_paper.pdf

starting at page 6..

That's very hard to read through. Call me skeptical. Red flags went up when I got to this sentence: These materials, being equivalent to the coating applied on the famed Stradivarius violins, help give the feeling of being present in a natural sound field and reproduce the true sound of musical instruments in higher fidelity. Stradivarius huh? So the varnish used on the wood in violins is really great to apply to paper speaker cones because it makes the reverb on your recording sound more natural. Who'd have thought it! I smell a business opportunity here. Violin varnish in tiny little bottles, call it TX-14 super-secret space-age damping compound, sell it for five hundred dollars a bottle, and you're done.

Plus, if that's the same woofer that they sell diy, the Xmax on those things looks pitiful. Can't imagine them going very loud or very low without problems. Speaking of, they actually took the NF1 off the American market, presumably because of problems related to an American habit of playing their monitors loud. Those nasty Americans and their loud music!

Heh. Sorry for any hostility there. Anyway, I think they're saying that stiffness and internal dampening are what's needed for a good speaker cone, and that their paper material is the best available compromise between the two. I didn't really understand the stuff regarding the surround construction.

Originally posted by ScottG note however that many poly-based loudspeakers are not "pure" poly designs that could contain:
1. structural differences (honycomb, woven, etc.)
2. coatings
3. core differences (sometimes graphite)
4. suspended substrates (sometimes graphite, or other carbon forms, aluminum, or magnesium)

I realize poly cones can be of more complex material combinations. I'm trying to figure out if there are poly-based drivers or speaker designs that can give the same level of detail that good stiff coned drivers have, such as the Accutons and metal Seas Excels. If there are, are these drivers just that well designed, or do they have that much better motors or what? I've managed to develop this predjudice against poly coned drivers somehow and am just trying to figure out if I was wrong. If I am wrong, I'd just as soon use some of the better poly drivers.

Ben.

 

[David Gatti]

Yeah. I've noticed that too. Seas doesn't yet have distortion graphs for these drivers like they do with the rest of the Excel line though.

Could you imagine the reactions if the paper cone produces clearly superior results to the ever-popular magnesiums?

 

[Salas]

In the 80's I remember an answer to a letter of mine from SEAS that besides asking for their new catalogue, I asked them if they were about to introduce new tech materials (back then Kevlar, Hexacone, TPX, Aerogel, Cobex, Yamaha Beryllium, were new stuff). They wrote me that hi-tec materials are not always the best bet.
Maybe this statement is still relevant.

 

[joe carrow]

Idea for modding Dayton RS drivers

Hey Noksukau,

Since you seem to have a background in materials- what do you think of this idea?

Bead-blast the surface of the larger Dayton Reference Series drivers to increase their stiffness by cold-working the outer surfaces, thus increasing the range before they break up.

Has anybody seen bead-blasting as a speaker mod for aluminum cones? It seems like it could work.

 

[5th element]

Could you imagine the reactions if the paper cone produces clearly superior results to the ever-popular magnesiums?

I think the important thing to remember here is the nextel coating applied to the paper cone. Nextel can be made VERY rigid and brittle as far as I know, and indeed the SEAS cones do have higher Q break up then most paper cones that we know of.

It would not be surprising that the new drivers do have lower distortion, not just because of the cone material but because of the motor. The mag excels have been out for a while now and presumably all the drivers use the same construction techniques and rough motor design.

These new drivers will have probably been in the R&D department for the better part of a year under going fine tuning to ensure (hopefully) better specs then the other drivers.

But we shall have to wait and see.

 

[Noksukau]

Re: Idea for modding Dayton RS drivers

Hey Noksukau,

Since you seem to have a background in materials- what do you think of this idea?

Bead-blast the surface of the larger Dayton Reference Series drivers to increase their stiffness by cold-working the outer surfaces, thus increasing the range before they break up.

Has anybody seen bead-blasting as a speaker mod for aluminum cones? It seems like it could work.

Interesting idea, and one which I'm still wrestling with the theoretical consequences of!

Firstly, cold-working will increase the yield strength of a material (by virtue of it already having yielded and so a higher stress is needed to reactivate or nucleate more dislocation sources), it will not however affect the stiffness to any similar degree. You see, yielding is a phenomenon associated with plastic (permanent) deformation whereas stiffness is an elastic property - determined primarily by the interatomic bond properties.

As an example of this, all steels have a Young's modulus of about 190-210GPa, despite varying massively in tensile strength. All the fancy alloying elements and heat treatments that go into making an extreme high tensile bolt compared to one of mild steel mean diddly squat when operating in an elastic environment. It just means that with the stronger bolt you remain in that safe elastic region for longer before those interatomic bonds give up the ghost and plastic deformation starts. Which in work is where everyone starts running!

Back to speaker cones though, you will not be permanently deforming the cone during normal use (I hope! ) so work hardening the surface will not affect its operation or properties. Surface treatments wouldn't contribute much to strength either, only scratch resistance. The bulk of the cone material would be unaffected and be the first weak spot for the onset of plastic deformation.

Similarly, cone break-up is partly dependent on those same interatomic bonds, like stiffness, and so that wouldn't be affected either. To be honest (having never been lectured in this sort of stuff) I'm not sure what the greatest contributor to break-up is, but natural material resonances will not be altered by increasing strength. For that you need to change the crystal structure by introducing new phases or changing material, but all that will do is shift your resonances somewhere else. With so many harmonics, you will still get a break-up that can't be internally damped (except maybe in a multi-phase alloy) and will feature somewhere in the audible range.

Therefore, I still think the best way forward for metal cones is to mix as many multi-phase materials as possible to emulate the spread spectrum of resonances that you get with paper. That way, when one particular alloy is in resonance, the others won't be and can damp its vibrations. A sandwich of many thin layers of spun foils of say, aluminium, magnesium, titanium etc. would work great I reckon. And even better, as I mentioned before, take two really thin cones made like this and join them with ribs or corrugations or a honeycomb structure, like an aircraft wing. Think of those old kit toy aircraft made of balsa wood with doped paper over the top, stiff and light.

Sorry to blow your shot-blasting suggestion out of the water, I like the idea though. I can just see a company picking up on it and claiming that it affects stiffness and break-up and surface wave propagation and all manner of other things that they could pass on to the marketing deparment! I'd patent it anyway.

Ben.

 

[5th element]

Re: Re: Idea for modding Dayton RS drivers

Therefore, I still think the best way forward for metal cones is to mix as many multi-phase materials as possible to emulate the spread spectrum of resonances that you get with paper. That way, when one particular alloy is in resonance, the others won't be and can damp its vibrations. A sandwich of many thin layers of spun foils of say, aluminium, magnesium, titanium etc. would work great I reckon.

Ben.

I can understand what you are saying here, but isnt the whole point of a metal cone to be free of resonance completely within the pass band? Surely adding in other alloys would bring us back to a controlled resonance situation such as paper and poly cones. Or are you just meaning reduce the resonance in a metal cone from say a 15-20dB peak to one that measures more like 5-10? This would have an advanted of reducing the level of the harmonics produced further down the frequency band.

 

[Noksukau]

Re: Re: Re: Idea for modding Dayton RS drivers

A bit of both I suppose!

Not wishing to sound like any kind of expert on this, I'm not, I still have a lot of learning to do and much of this is just me trying to apply basic (and assumed) theories to what is arguably a rather complex scenario.

I can understand what you are saying here, but isnt the whole point of a metal cone to be free of resonance completely within the pass band?

As well as being stiffer, yes. However on many speakers that pass band ends up being rather narrow due to having to account for harmonics lower down the spectrum associated with the break-up nodes (as Pan was arguing for in the "Intelligibility of voice" thread recently). Either that or you live with the 'metallic' sound those resonances impart.

Surely adding in other alloys would bring us back to a controlled resonance situation such as paper and poly cones.

Yes, but surely resonance control is a good thing. Paper's heterogenous nature is what gave me the idea for combining alloys. Bear in mind that this damping will only be occurring where any of the individual metals would be resonating anyway, so whatever usable pass-band you might have had before can be extended. I don't know what kind of resonances typical paper cones have within their pass-band (or rather, usable range without break-up effects, as you can put the pass-band where you like ), you seem to be implying that they do have some, and metal cones don't? I don't know. I would have thought paper was free from resonance in its usable range, but then you have the problem that its usable range can be defined as that portion without resonance! Arrrggggh!

Anyway, the idea was to control the metal's resonances in the same manner that paper is self-damping, but that wouldn't rob the metal of its naturally resonant-free pass-band that you mentioned. Best of both worlds, in theory.

Or are you just meaning reduce the resonance in a metal cone from say a 15-20dB peak to one that measures more like 5-10? This would have an advanted of reducing the level of the harmonics produced further down the frequency band.

Not pretending that resonances could be eliminated, yes, it would hopefully have the effect of damping those resonances. And as you say, this would reduce those lower frequency harmonics, and delay the main break-up peaks, allowing you to use the woofer over a broader frequency range, increasing the pass-band and generally getting closer to an ideal piston.

 

[joe carrow]

Re: Re: Idea for modding Dayton RS drivers

Sorry to blow your shot-blasting suggestion out of the water, I like the idea though.

Ben.

Hey, no sweat! Now that you mention it, I recall a time in school when I made this error- it only feels bad because I made the same mistake twice. It was cool to get a good reply from you.

So, this is getting pretty far off topic- but do you suppose that you could stiffen a cone through one of these methods diamond vapor deposition, or perhaps silicon carbide- like they use to treat machine tools?

Really, the idea here is to push resonances farther out of the passband, and if possible above audible perception. If they're overdamped, that works too. I hear that one of the problems when working with the Seas magnesium drivers is that the second and third harmonic of a tone can be amplified when it excites a breakup frequency.

 

[454Casull]

The diamond deposition idea is workable, although there is a size limitation. Also, diamond is rather brittle and may crack after use.

 

[Joules]

It' about the sound speed of the material.

When the voice coil first trys to move the outeredges of the cone don't know anything about it until the impulse traveles out to that part of the cone. So the voice coil and former start moving before the outer part of the cone. The higher the freq. and the larger the cone and the softer the material (slower the sound speed of the material) the worse the problem gets. This traveling wave reflects off the outer edge and starts traveling back to the voice coil and now starts to creat standing waves. altho aluminum has a high sound speed it can't damp the standing waves very well. Paper and plastic have a much slower sound speed but can damp the standing waves better.