AudioRoundTable.com: Pi Speakers

Home » Sponsored » Pi Speakers » 2500 hz hump

Show: Today's Messages :: Navigator :: Polls
E-mail to friend

2500 hz hump [message #48390]

Tue, 13 December 2005 16:33

Paul C.
Messages: 218
Registered: May 2009

Master

Wayne, besides the woofers you use in Pi's, I have been looking at other similar high efficiency woofer. It seems that most have a hump or peak in the 2000 hz range, then fall off. Examples are Eminence Gamma 12,

Some the hump is quite mild, Eminence Delta 15LF, Delta 12LF, Gamma 15.

What causes this peak? Is there any way to tame this in the crossover?

Report message to a moderator

Re: 2500 hz hump [message #48391 is a reply to message #48390]

Wed, 14 December 2005 08:25

Wayne Parham
Messages: 18787
Registered: January 2001

Illuminati (33rd Degree)

There are a couple of things that cause midrange response to rise in a midwoofer. One is collapsing directivity and the other is cone flex resonance, modes where the cone is twisting and a resonance sets up on the membrane. Collapsing DI causes a flat line of rising response but resonance modes cause peaks. You can equalize collapsing DI in the crossover, and it is a useful feature when paired with a HF horn, because you can crossover in the frequency range where directivity is similar between the two subsystems. Breakup modes are a little trickier to work with, because they make the driver surface act a sort of like a chaotic array of sound sources that change with respect to frequency.

Report message to a moderator

Re: 2500 hz hump [message #48392 is a reply to message #48391]

Wed, 14 December 2005 08:43

Paul C.
Messages: 218
Registered: May 2009

Master

Wayne, I am sure you have probably discussed this before, but thanks for the review!

Report message to a moderator

Re: 2500 hz hump [message #48393 is a reply to message #48391]

Wed, 14 December 2005 11:49

Spinjack
Messages: 100
Registered: May 2009

Viscount

The modes are the reason that some high-end speakers use drivers with random, but intentional, "irregularities" in the cone, is it not? For instance, Sonus Faber has those epoxy filled slits that breakup the surface. Or am I off in left field on this one?

Report message to a moderator

Re: 2500 hz hump [message #48394 is a reply to message #48393]

Wed, 14 December 2005 14:10

Wayne Parham
Messages: 18787
Registered: January 2001

Illuminati (33rd Degree)

There are lots of things manufacturers have done to reduce the problems surrounding breakup modes, mostly by working with different materials and cone geometries.

Report message to a moderator

Re: 2500 hz hump [message #48395 is a reply to message #48394]

Wed, 14 December 2005 17:04

Matts
Messages: 359
Registered: May 2009

Grand Master

by nature of being woofers, arent' these drivers made to cross below this point? If the mfr's are trying to optimize performance in a given waveband, there are bound to be trade-offs outside of that waveband. And, if a mfr tries to make a single-driver solution, there are trade-offs in that, too (called "BASS"). No free lunches!

Report message to a moderator

Re: BTW.... [message #48396 is a reply to message #48395]

Wed, 14 December 2005 17:05

Matts
Messages: 359
Registered: May 2009

Grand Master

the Omega 15's in the Stage series, besides sounding good, have a very pretty freq. res. graph.....

Report message to a moderator

Re: BTW.... [message #48398 is a reply to message #48396]

Wed, 14 December 2005 20:04

Paul C.
Messages: 218
Registered: May 2009

Master

Yes, they should cross over before that 2000 hz-2500 hump. But even with the crossover rolling them off, they still may have a high peak that interferes with getting an accurate xover.

With some, even with 12 db/oct, this peak makes them unsuitable for the 1600 hz xover freq recommended for the Eminence PSD2002. A 2-way will not work for these, you would have to go much lower, and probably go to a 3-way system.

Report message to a moderator

Re: BTW.... [message #48401 is a reply to message #48398]

Thu, 15 December 2005 07:57

Spinjack
Messages: 100
Registered: May 2009

Viscount

Can you not just set the crossover point such that the signal rolls off just before the hump which combined with the hump gives relatively flat total response until the back side of the hump where the total rolloff is now steeper than 12db/oct? You would have to design the system for a steeper rolloff total on the back side of the hump.

(Forgive me is I'm way off base. I'm still rather new at this.)

Report message to a moderator

bad breakup [message #48409 is a reply to message #48401]

Fri, 16 December 2005 22:41

ToFo
Messages: 219
Registered: May 2009

Master

if it's just rising response from the collapsing directivity of the large driver then you are right on. you can set the crossover electrically lower than your real acoustic target. You will not get an ideal summation of the two drivers, but you can take steps to insure you don't cancel out either (or just hook the tweeter up in and out of phase and see which is louder at your crossover frequency, and hope for the best). A calculated second best is not necesarily a bad thing.

Cone breakup resonances are another matter altogether. I believe all the big drivers have some past 800 Hz or thereabouts, so it's more a matter of how bad are they and which sound least offensive. There can be a lot of sound still coming off the cone after the event is over because the cone is ringing like a bell at certain frequencies. You can cross it over or EQ it all you want, but it can still quack like a duck if the driver is spikey up high. I heard a frat party system once that had a hot 18 crossed to a small 3500 hz horn. Didn't matter what the instrument was, handclap - quack!, Snare - Quack!!, saxophone - QUACCKKK!!, you get my drift. Most frequency plots are set so the "pen" moves too slow to show how bad breakup really is, so you just have to be a carefull shopper and find designers/posters you trust (guess your in the right place already).