soft vs hard mid-woofers cones: audible directivity characteristics?

[crazyhub]

Hi all,

During the design of my last two-way using a soft material mid-woofer (SEAS U18), Alex, one friend of mine, simply replaced the U18 with a L18 (alu hard cone); filter was modified as to obtain the same (on axis) fr response, same acoustic LR4 Fx_2kHz; comparison was instant by using the crossover emulator of LSPcad. An obvious high-mid clarity using the L18 was the result of this quick test.

Alex thinks that one of the advantages of hard vs soft is the higher membrane break-up which allows better directivity pattern thus better in-room power response and for him this was the main reason of what we heard.
Comparing these 2 drivers on SEAS graphs, the measured difference is not obvious to me:
THE ART OF SOUND PERFECTION BY SEAS - H1571-08 U18RNX/P
THE ART OF SOUND PERFECTION BY SEAS - H1224-08 L18RNX/P
At the moment I don't find out the reason of this strange difference (harmonic distorsion is low enough on both drivers so I thought it should not be audible at average listening level)...
Any ideas guys?

[jerome69]

Hi Hubert,

Directivity is not the explanation of this difference of sound. A directive driver doesn't mean less clarity. Directivity is essentially a function of size : the diameter, geometry of the driver and the size of the baffle. Imagine the sound of a 5" and a 8" with the same cone, the 5" will sound more clear because it's less directive ?

This is the difference between high end drivers and normal drivers. Better "clarity" in the operating area. It 's rather a function of the motor and the cone.
In your case it simply means than the aluminium cone has a better non linear distortion than than the woven polypropylene cone. To test this you use Linkwitz distortion test with burst wave. A complex test very close the way we listen music. With normal test like Zaph's tests you don't see anything.
See:
Midrange distortion test
SL's test of NEO3

Cheers.

[6.283]

Hi Crazyhub,

check the upper right corner (Shrinking Radiation Area?) of the poster from Klippel.
http://www.klippel.de/ddl/files/Post...ter_Page_1.pdf

That explains the directional behavior of "soft" vs. "rigid" cones. I don't think that this is what made the difference in your case as it would only ocurr above 2KHz with an 6.5" driver.

[David Gatti]

Using a 7" poly cone up to 2kHz will always cause a loss of detail compared to an aluminium cone. The CSD plots will shed a lot of light on this phenomenon. Check out Zaph Audio's charts for a comparison of CSD behaviour.

[LineArray]

Originally Posted by crazyhub

Hi all,

During the design of my last two-way using a soft material mid-woofer (SEAS U18), Alex, one friend of mine, simply replaced the U18 with a L18 (alu hard cone); filter was modified as to obtain the same (on axis) fr response

Hello crazyhub,

the question is how similar is "same" in this example.

In the manufacturers sheets the ALU version seems somewhat pronounced
around 850 Hz which occurs even under angles and might also show up
in the power response.

A slightly more "present" quality of the sound could be the result and
especially voices sounding different.

A waterfall spectrum of both speakers would be interesting too.

If you take the "soft cone" version and use a parametric EQ, you might
try, whether you can imitate the sound of the AL Version to some
degree by making it slightly hot around 1Khz.

If differences in decay are the main reason, you will not succeed in
making both speakers similar by EQing.

I know of some AL cones which have very short decay throughout the
midrange and up to the lower treble.

[alspe]

crazuhub, I don't see how cone breakup could allow better directivity pattern?


Stiffer cone acts more like piston than softer one. That allows better transients but might cause ringing.

[crazyhub]

to 6.283:
thanks for the link to Klippels file. I tend to agree with you that in the presented case the audible differencies aren't due to directivity behaviour.

to alspe:
Klippels link demonstrates that stiffer cones allow to minimise directivity because it act purely pistonic higher in fr.

to David:
In Zaphs charts the CSD plots are rather in favour of the U18...

to Jerome:
Except a H2 problem close to 700Hz, also seen on the Zaphs fr and impedance plots (but less in my own U18), the distorsion is of same amplitude for both drivers; moreover the U18 doesn't suffer from odd orders distorsions in the upper-mid du to the huge break-up of the alu version. My 2-way distorsion 90dbs/1meter:



to LineArray:
I agree: similar fr response isn't identical...
<<< A slightly more "present" quality of the sound could be the result (...of the pronunced 850Hz area) and especially voices sounding different.
*** I said improved "clarity" but enhanced "warmth" is a better world. "Warmth" doesn't come from upper-mid, right? rather from mid/low-mid. Unfortunately I'm not able to continue comparisons or measurements or EQuing this speaker...
If my memory is good, we tried to EQ the U18 but didn't succeed in giving it a similar sound, so I'm pretty sure the difference was intrinsically due to the L18 properties.

Thanks guys for your replies!

[MisterTwister]

Aluminum cone ringing is perceived as high-mid clarity. I would not cross this driver anywhere above 1khz. Check Zaphs distortion graphs, that 7khz ring is still visible around 1.5khz. Paper cones are just as bad, except their ringing sounds different.

[alspe]

Seems that stiffer cones are good if they are in multi way (more than 2) speaker and crossed higher than 2nd order slope. For example L22 for bass, L15 for mids + tweeter, 4th order slopes. And where cones are driven harder.

But how much worse distortion behaviour of softer cones matters in real world IF we can make speaker bigger, more sensitive? For example WTW with U/CA18 soft cone versus WT with L18. Former is 4-6 dB more sensitive.

[6.283]

Originally Posted by MisterTwister

Paper cones are just as bad, except their ringing sounds different.

That is over generalized and would also depend on the diameter of the driver.
Those things have to be evaluated for each and every driver that is a potential candidate for a design.