View Full Version : 2382a and 2450 combination?

01-26-2011, 03:24 AM
Dear all,

Is there anyone who may have measurement curves for the 2382a horn together with the 2450H (or J) driver. I seem to get a nasty resonance at 15kHz. Please see attached MLS measurement.

Donít really know why. It is the same for both left and right speaker. It is a far field measurement at 63cm, XO is LR @ 1,2 kHz and the mic is front center. Mic is a calibrated M30 so that is not the problem and the time window is only 3ms so there should be no reflections distorting.

Can it be that this combination does not work well? The curve looks pretty much like the referens curve in the 2450 spec sheet when mounted on plane-wave terminated tube! The shifts are extra visible due to the scale.

Also does anyone have a setting for an electronic (digital) filter for this combination. I have searched the Internet and most of the JBL websites for a setting but without luck. There should have been someone using this combination with the DCS280 or similar. I have not found any settings for 2382 and 2446 either.

All input appreciated

Best regards

01-27-2011, 08:29 AM
The NON-Beryllium 4" diaphragms all produce serious resonances above 9-10 kHz. Normal.

Can you please plot that diagram with a 5db/division scale (e.g. +10db to -40db). It will look much nicer and will be comparable to other diagrams.

01-28-2011, 09:30 AM

Attached is a new measurement with the requested scale. Same resonance....:crying:

Best regards

01-28-2011, 11:03 AM
Your latest response curves are very similar to the ones I am getting with a 2382a/2446h combination.
In the case of the 2446h driver, I am seeing the same abrupt roll-off above 8 kHz and the same choppy response above 10 kHz.
And I donít recall seeing as big a spike at 15 kHz as you are getting with 2450.
But, the resonance is still pretty rough up there.

Have you considered adding 2405H UHF drivers to your project?
I realize that it is an expensive solution to your problem.
But, even with a lot of EQ the 2382a horn and a 4" diaphragm is not going to make it out to 20 kHz very easily.
In my own system, I use an Ashly 4001 electronic crossover to feed the 2382a/2446h with a frequency range of 1200 Hz to 8.5 kHz.
And then use a 2405 driver to reproduce everything above 8.5 kHz.
And to compensate for the very slight roll-off just below 8.5 kHz, I added a simple passive RC shelving network between the HF amplifier and the 2446H driver.
The network adds about 5 db of boost to the 2446h driver at 8.5 kHz.
Without, this shelving network, I noticed a dip in the system response just below the 8.5 kHz crossover point.

Alternatively, you could switch to a more directional horn like the 2380 or 2385 as these horns tend to have a higher roll-off point.


01-28-2011, 11:43 AM
The aberrations in the phase response are what's telling. The driver will not be made to behave up there, you can add energy to the driver and get more energy out and it will show up in the magnitude response but it will be noise.

A large part of why these sound so much better with a real tweeter is getting that noise out of the system by creating the energy with something that can do the job.