What is generally the best material for speaker cones with the least compromise and most benefits?

[ahofer]

Take a look at the W18 graphs above, near 5kHz. Obviously, those rather stiff cones deform rather badly up there.

Which? is RADIAL measured somewhere upthread? I didn’t see it. That’s a lot higher than they cross them over, of course.

 

[Ron Texas]

Unobtanium.

 

[ahofer]

Unobtanium.

See #39

 

[Ron Texas]

See #39

That's what I get for not reading the entire thread.

My second suggestion is silly putty.

 

[top gnome]

Hi new member here. Years ago I built the thors with the magnesium drivers. I would warn if you live close to the ocean the drivers may corrode mine did to the point were they perforated. Seas and Solen thought the nextels may be a better fit. The nextels are good drivers and they sound almost as good as the mag drivers but they do lack some of the lowest bass. I was able to use the same cross with minor changes and the measurements are very similar within a couple of db.

 

[fpitas]

Which? is RADIAL measured somewhere upthread? I didn’t see it. That’s a lot higher than they cross them over, of course.

Post #45 has the W18 response graph about halfway down. You can still hear the peak, even when it's in the stopband. Generally it's good to push it down by 40dB from nominal.

 

[LesterNZ]

Fan-boy here for using Be tweets & mids. IMHO, it's mostly about the impulse response showing nearly no ringing on the lagging edge transient. No stored - delayed energy. Just the electromagnetic to barometric. Some details here:

https://materion.com/-/media/files/beryllium/engineering-design/acoustic-properties-of-beryllium.pdf

My own ScanSpeak D3004/664000 Be's as measured by Troels

SS-domes

Heard some Yamaha NS-1000's ages ago whilst being the sound-op in a live organ trio recording at Blues Alley Jazz in Wash D.C. In the "same room" those monitors set-up for the master tape play-back at evenings end had the best clean, clear reproduction of tapped cymbals, brushes, snare you could ever wish for. The drum kit was right there for immediate A/B and the only problem in this ideal comparison was limited dynamic headroom from the McIntosh 275 not driving realistic levels into those Yammie's. The Hammond/Leslie B-3 pedal was simply too much, though the tenor sax and "non-bashed" percussion was very much 'simply superb reproduction'. (Couldn't afford them back then, though I should have sold the car, the girlfriend, the Small Maggies, (or the Picasso, Manet diamonds, rubies etc too!)).

I have replaced the co-axial mid-hf aluminum dome drivers in my Spatial M3-T double 15's open baffles with the ScanSpeak D3004/664000. XO'd at a low 850 Hz via L-R 48dB/oct in a miniDSP SHD to Kinki MX-7 (class AB) power amp. And, at reasonable volumes, my 'less than optimum' throat coupling-loading is inconsequential thanks to tweaking & smoothing from the SHD's excellent EQ & DIRAC flexibility.

It's an absolute treat to hear the minor and major sonic differences in instruments and recording techniques while being functionally, yet gently, pulled right smack into that reproduced club, studio, or concert session. Listening ecstasy.
J

 

[Barrelhouse Solly]

I dunno. This seems to be in the same category as "What is the best guitar tone wood?"

 

[Salt]

Purely personal perception:
there is no ideal oder optimum in material.
A diaphragm/menbrane/conus/whatsoever should provide minimum distortions and losses within the range it is intended to work, far off at least 1 octave (better 2) to distortion peaks.

 

[MattHooper]

SEAS Graphene Coated Magnesium drivers, of course. Just because my speakers use 'em ....and they look cool...

But when I feel like listening to aluminum for a midrange, I prefer "ribbed" for pleasure
(Thiel 2.7s...)

 

[mhardy6647]

That's what I get for not reading the entire thread.

My second suggestion is silly putty.

That'd be excellent -- we could emblazon our drivers' cones with images of our favorite musicians -- as long as we could find some newsprint to press against the cones.

 

[MAB]

SEAS Graphene Coated Magnesium drivers, of course. Just because my speakers use 'em ....and they look cool...

I do like those Seas drivers. They are an interesting case study to the OP's post: find the best cone material, minimizing compromises. Those high-tech Seas metal alloy cones are not quite Beryllium or Unobtanium, but close. A by-product of the low mass and high stiffness of the cones is a large breakup mode. Jeff Joseph actually mentions this in a comment in the Stereophile review. I have some Seas Excel drivers somewhat similar to the Joseph Audio Perspective, I measured them since they are very interesting drivers:

Here is the Seas spec sheet and and a measurement I made for comparison (ignore the lower frequencies due to baffle vs. box):

You can see why I include the manufacturer's data just because the peak is so dramatic. I can measure the same spray of harmonics. I can tell you it sounds horrible if you do a frequency sweep with no filter.

There is actually lots going on, peaks and valleys are a result of the cone both turning into a sombrero shape, and also potato-chip shape. For instance Robert-H Munnig Schmidt with RMS Acoustics & Mechatronics / Grimm Audio in this paper:

These distortions of the shape of the cone have harmonics around the fundamental breakup mode.
They can easily be measured too, frequency response at top and distortion below:

Interesting, the dominant distortion peak isn't the fundamental, its a third-order HD peak at 1.72kHz, exactly 1/3 of the breakup mode! The 5th order peak is also poking it's head. Odd HD is pronounced. This doesn't seem good! While there are ways to mitigate the resonance, it might be good to see how this compares to other drivers with different cone materials. Here are four; two metal woofers from Seas, a B&C midrange made from coated woven fiber, and a paper-cone JBL 15" woofer that can play fairly high frequencies.

I measured the distortion of all four drivers at varying levels, selected some relatively reasonably loud samples of each driver where I wasn't clipping a speaker of mic too badly, all roughly level-matched at 1kHz:

The distortion of the Seas W18 is actually quite under control below 1kHz, much better than the 5" midrange! So is the 10" Seas W26! The JBL 2216-Nd actually is even lower distortion. This is in some ways unfair, since the 2216 has so much output capability. Here is the JBL at successively higher SPL getting quite loud.

While the % distortion increases, it isn't driven by materials so much as application. Drive these alloy cone woofers outside of the useable band and they sound horrible. The ringing and distortions are quite harsh sounding, and have show up dramatically in the measurements. Even simple filters are not enough to deal with these drivers, and advance methods need to be applied to get the best result.

In this apples to oranges comparison, we see the big 15" paper cone JBL is actually the lowest distortion, with a very useable response to above 1kHz, with minimal peaks and valleys and distortion under control. You can essentially use it at any frequency it has output.

It would be good to redo this with drivers with the same size, but different materials. For sure the exotic drivers need special care, and limit the choices in crossover slopes and frequencies, and even then might need notches and other approaches to make them sound good.

I didn't answer OP's question...
I am pretty sure there isn't a clear answer.

 

[Pearljam5000]

Whatever Genelec is using...
Kidding
I'd say for the tweeter , Beryllium is the best.

 

[fpitas]

Interesting, the dominant distortion peak isn't the fundamental, its a third-order HD peak at 1.72kHz, exactly 1/3 of the breakup mode! The 5th order peak is also poking it's head. Odd HD is pronounced. This doesn't seem good!

Bruno showed that the best way to deal with the peak is a parallel trap leading to the woofer, to choke off the current, versus a series tuned circuit to ground to shunt the driver. The distortion is significantly better. I'll try to find the paper he wrote for Purifi.

 

[fpitas]

Low Distortion Filter for PTT6.5X04-NAA

purifi-audio.com

 

[MattHooper]

I do like those Seas drivers. They are an interesting case study to the OP's post: find the best cone material, minimizing compromises. Those high-tech Seas metal alloy cones are not quite Beryllium or Unobtanium, but close. A by-product of the low mass and high stiffness of the cones is a large breakup mode. Jeff Joseph actually mentions this in a comment in the Stereophile review. I have some Seas Excel drivers somewhat similar to the Joseph Audio Perspective, I measured them since they are very interesting drivers:

Here is the Seas spec sheet and and a measurement I made for comparison (ignore the lower frequencies due to baffle vs. box):
View attachment 333252

You can see why I include the manufacturer's data just because the peak is so dramatic. I can measure the same spray of harmonics. I can tell you it sounds horrible if you do a frequency sweep with no filter.

There is actually lots going on, peaks and valleys are a result of the cone both turning into a sombrero shape, and also potato-chip shape. For instance Robert-H Munnig Schmidt with RMS Acoustics & Mechatronics / Grimm Audio in this paper:

View attachment 333254
These distortions of the shape of the cone have harmonics around the fundamental breakup mode.
They can easily be measured too, frequency response at top and distortion below:
View attachment 333260
Interesting, the dominant distortion peak isn't the fundamental, its a third-order HD peak at 1.72kHz, exactly 1/3 of the breakup mode! The 5th order peak is also poking it's head. Odd HD is pronounced. This doesn't seem good! While there are ways to mitigate the resonance, it might be good to see how this compares to other drivers with different cone materials. Here are four; two metal woofers from Seas, a B&C midrange made from coated woven fiber, and a paper-cone JBL 15" woofer that can play fairly high frequencies.
View attachment 333261

I measured the distortion of all four drivers at varying levels, selected some relatively reasonably loud samples of each driver where I wasn't clipping a speaker of mic too badly, all roughly level-matched at 1kHz:
View attachment 333274
The distortion of the Seas W18 is actually quite under control below 1kHz, much better than the 5" midrange! So is the 10" Seas W26! The JBL 2216-Nd actually is even lower distortion. This is in some ways unfair, since the 2216 has so much output capability. Here is the JBL at successively higher SPL getting quite loud.
View attachment 333280
While the % distortion increases, it isn't driven by materials so much as application. Drive these alloy cone woofers outside of the useable band and they sound horrible. The ringing and distortions are quite harsh sounding, and have show up dramatically in the measurements. Even simple filters are not enough to deal with these drivers, and advance methods need to be applied to get the best result.

In this apples to oranges comparison, we see the big 15" paper cone JBL is actually the lowest distortion, with a very useable response to above 1kHz, with minimal peaks and valleys and distortion under control. You can essentially use it at any frequency it has output.

It would be good to redo this with drivers with the same size, but different materials. For sure the exotic drivers need special care, and limit the choices in crossover slopes and frequencies, and even then might need notches and other approaches to make them sound good.

I didn't answer OP's question...
I am pretty sure there isn't a clear answer.

Thanks that is a very interesting analysis...and a prime example of why this forum can be so great!

I see you measured some SEAS drivers different from the ones used in my particular Joseph Perspectives Graphene 2 speakers. Outside what JA provided in the stereophile review, I've never been able to confidently find measurements of the drivers used in my speakers. The spec mentioned for my speaker's drivers is "two 5.5" (140mm) magnesium-cone woofers." Early on I found that someone had measured the "new" SEAS graphene drivers, though the size is referenced as "SEAS Excel Graphene...5.25" Driver. So I'm not sure if it's the same driver. Anyway, the measurements are similar to what you got:

Test Bench: SEAS Excel Graphene W16NX003 5.25” Midbass Driver

In this Test Bench, Vance Dickason characterizes the SEAS W16NX003 5.25” midbass driver, one of the new graphene-coated magnesium cone drivers from the brand's Excel Graphene range. This product line has been an ongoing project for more than two years, combining the already highly prized...

audioxpress.com

The writer seemed fairly impressed all things considered.

As to break up modes, I never care much about that since I only care about how things sound in the end product. If the break up is dealt with well enough, it's no issue.
Joseph Audio was originally known for using Richard Modafferi's "infinite slope crossover," which I'm guessing you'd know more about than I do, but is a purportedly super steep crossover (purportedly dropping 120dB per octave for the first 42dB). Which if correct would seem to help realize the potential of many metal drivers. I understand the slope is somewhat controversial, and also Jeff Joseph has modified the crossover through time (I think he's crossing over the tweeter lower these days too). Further, I don't think the stereophile measurements showed anything particularly heroic was going on with the crossover?

In any case, it seems to me Joseph knows what he is doing in terms of the final result, because despite the nasty break up stuff in these drivers, the end result is super smooth sound. That is the specific feature of these speakers that is remarked upon over and over by listeners at audio shows, reviews, and by owners. The smoothness and apparent lack of distortion is the very first impression they made on me, and continue to make on me.

Cheers.

 

[617]

Thanks that is a very interesting analysis...and a prime example of why this forum can be so great!

I see you measured some SEAS drivers different from the ones used in my particular Joseph Perspectives Graphene 2 speakers. Outside what JA provided in the stereophile review, I've never been able to confidently find measurements of the drivers used in my speakers. The spec mentioned for my speaker's drivers is "two 5.5" (140mm) magnesium-cone woofers." Early on I found that someone had measured the "new" SEAS graphene drivers, though the size is referenced as "SEAS Excel Graphene...5.25" Driver. So I'm not sure if it's the same driver. Anyway, the measurements are similar to what you got:

Test Bench: SEAS Excel Graphene W16NX003 5.25” Midbass Driver

In this Test Bench, Vance Dickason characterizes the SEAS W16NX003 5.25” midbass driver, one of the new graphene-coated magnesium cone drivers from the brand's Excel Graphene range. This product line has been an ongoing project for more than two years, combining the already highly prized...

audioxpress.com

The writer seemed fairly impressed all things considered.

As to break up modes, I never care much about that since I only care about how things sound in the end product. If the break up is dealt with well enough, it's no issue.
Joseph Audio was originally known for using Richard Modafferi's "infinite slope crossover," which I'm guessing you'd know more about than I do, but is a purportedly super steep crossover (purportedly dropping 120dB per octave for the first 42dB). Which if correct would seem to help realize the potential of many metal drivers. I understand the slope is somewhat controversial, and also Jeff Joseph has modified the crossover through time (I think he's crossing over the tweeter lower these days too). Further, I don't think the stereophile measurements showed anything particularly heroic was going on with the crossover?

In any case, it seems to me Joseph knows what he is doing in terms of the final result, because despite the nasty break up stuff in these drivers, the end result is super smooth sound. That is the specific feature of these speakers that is remarked upon over and over by listeners at audio shows, reviews, and by owners. The smoothness and apparent lack of distortion is the very first impression they made on me, and continue to make on me.

Cheers.

People seem to really like the Joseph Audio stuff. An incredibly steep filter necessitates close driver sizes, which they don't seem to use in their designs - I think they're using the 15-17cm units with ~28mm tweeters. This wouldn't be a problem with a LR2 filter, but with 'infinite slope' I'd be looking for a truly small midrange.

Does Joseph use the 12cm midranges in any of their speakers? I do like how small and precise looking they are, they always struck me as very high quality designs.

 

[ryanosaur]

SEAS Graphene Coated Magnesium drivers, of course. Just because my speakers use 'em ....and they look cool...

Ooh! That Seas Millennium Tweet is a pretty impressive Driver too! Is that part of your Speaker, as well?

 

[DonR]

Kevlar because it makes the driver "bulletproof"... and because that is what my speakers have.

 

[MattHooper]

Ooh! That Seas Millennium Tweet is a pretty impressive Driver too! Is that part of your Speaker, as well?

Yes.

I don't know about the technical behaviour of that tweeter, but man is it sweet and pure sounding.

I had an audio buddy and reviewer over at my place recently. He listened to a lot of tracks on my system including ones he knows really well. When I asked him to describe the sound afterward he said, aside from the crazy imaging, the thing that stuck out was the smoothness of the sound. In particular he said the highs didn't sound remotely rolled off, but really extended, airy, cymbals and everything else had wonderful shimmer, but it was "buttery smooth" and super relaxing at the same time, like he could just keep turning up the sound with no sense of fatigue. Which is exactly what I hear, and one of the main reasons I bought the speakers (which is very helpful because I suffer from tinnitus and hyperacusis, and so a speaker has to be really smooth in order for me to listen to it loud).