Converting 2445 to wider range more extended HF? [Archive]

jerv

12-21-2009, 12:31 PM

Extending the 2445's response to ~20kHz is possible, but at the cost of about 10-12dB sensitivity.

FWIW, according to my measurements, the 2445 (with stock diaphragm in a 2360 horn) drop off from 1 kHz at a slope of about 5dB to 5kHz. Then it drops off somewhat steeper up to 10kHz, which is about 12dB down, again compared with 1 kHz. Between 10 and 20 kHz there appears to be two resonances, extending the response somewhat. See attached graph, black curve.

If you attenuate the response 800Hz-7kHz enough, the 2445's response is extended upwards to 20 kHz. I have made a passive equalizing network for the 2445/2360 combination with intended crossover at 600 Hz and response from 800 Hz to 20kHz within +/- 2.5dB. To achieve this, the 1kHz level must be damped about 12 dB. See attached graph, blue curve.

With a supertweeter (like the 2405 or 2404) crossed in at about 9 kHz, the response does not have to be attenuated nearly as much. See attached graph, red curve.

The actual levels and crossover values may differ, but in principle the same is achieveable in other horns: like the 2380.

Allanvh5150

12-23-2009, 02:40 AM

If you are running a 2445 in any system it will have to be attenuated to match the speaker that you put under it. 14db or so with a 2225 and 18db or so with a 2235 so it will not be a problem.

Allan.

pos

12-23-2009, 03:10 AM

Mike Caldwell

12-23-2009, 07:18 AM

The diaphragms are interchangeable so you could try a 2446, 2450 or 2450-SL diaphragm in the 2445.
What type of horn are you using. Many "CD" constant directivity horns require the use of high frequency boost applied to the driver for the horns to operate as designed. The applied EQ is generally called CD eq or CD boost and starts about at 4,000hz with a 6 db per octave boost out to 20,000hz. The JBL 2385 horns in my main PA cabinets need such EQ.

jerv

12-23-2009, 09:14 AM

The difficulty will be to find a 2" horn with an acceptable UHF directivity.
Having energy on axis will not be very usefull if you get nothing off axis, like with an 2311.
Even the 2360 with its narrow slot is beaming above 10khz in the horizontal plan. And the vertial is much worse and show you what you will get with any 2" horn that does not have a diffraction slot.
http://www.jblpro.com/pages/pub/components/23606566.pdf

But the having a wide UHF dispersion will reduce the on axis UHF output...
I think you will be better off using a tweeter above 10khz

This is very well put. I agree completely. My own listening test confirms this: even if the frequency response measures ok on axis, the listening impression is one of a slight "dullness". In my system, I use a 2405 from 9kHz.

But, as to the OP's opiginal question: it really is possible to extend the 2445 to 20 kHz - even if it sounds better with a tweeter.

4313B

12-23-2009, 09:17 AM

But, as to the OP's opiginal question: it really is possible to extend the 2445 to 20 kHz - even if it sounds better with a tweeter.Even the 476Be sounds better with an 045Be. :D

cosmos

12-23-2009, 09:02 PM

Thanks for all of the replies.

Yes, I am trying to get to a decent, highly efficient 2 way. I would like to use a larger (1.5" or 2") comp driver because of their dynamics and fullness..

Presently, I have 2445 running off of a single ended tube amp on an electronic crossover from 600Hz up on 350Hz Edgarhorns. I am giving them a 12db/oct boost starting at about 8KHz with the EQ in the crossover. I have replaced the rotten foam in the back cap of the 2445s with 1/2" think heavy felt sound deadening material. This seems to have helped the natural tendency of the driver to sound a little aggressive. The 600Hz down band is being handled by 2226 drivers in 4507 cabinets. I am not at all unhappy with the sound of this system, as it is, as long as I am in the very small laser like zone. The system, when doctored up with EQ, sounds good and looks very good on RTA. For some reason, yet to be determined, I have a significant dip at 250-350 Hz. So far I have put a parametric boost on it to address it even though I can't reason out the source.

However, as good as it sounds in the zone, I feel really constrained by the Edgarhorns laser like beam. I know I will be giving up imaging, but I want to go to a CD waveguide to increase dispersion. Further, I know I will be missing even more energy (8KHz plus) off axis which is why I am trying to extend response, as much as I can naturally. I fully expect to have to equalize it to get there. Also, I would like to smooth the aggressiveness of the titanium.

Once I settle on the right Compression driver and waveguide, I intend to build WAF approved cabinet(s).

Sound logical?

Mr. Widget

12-23-2009, 09:16 PM

But, as to the OP's opiginal question: it really is possible to extend the 2445 to 20 kHz - even if it sounds better with a tweeter.:yes:

I haven't actually tried it myself with 2445s, but I did use 2397 horns and TAD 4001 drivers... supposedly good to 20KHz. When I EQ'd them to get them really flat out to 20KHz, I thought they sounded a bit harsh.

One of our two-way promoters... Zilch has tweaked numerous JBL and other drivers to measure quite nicely out to 20KHz on a variety of JBL horns... virtually every one of them sounded too harsh for my taste as well.

I'd love to find a horn/driver combo that could work as a nice two-way without the tweeter, but so far I have been disappointed. I have heard the Everest II with the tweeter blocked and it did sound quite nice, but unfortunately virtually none of us can afford them.

Widget

Mr. Widget

12-23-2009, 09:20 PM

However, as good as it sounds in the zone, I feel really constrained by the Edgarhorns laser like beam. I know I will be giving up imaging...Yep... those Edgar horns do image remarkably. I'd hate to give that up, but the head in the vice effect is pretty bad.

Good luck finding just the right compromise. There is no way you will find a perfect two-way with excellent imaging, but with luck you may get acceptably close.

Widget

Zilch

12-23-2009, 10:26 PM

I'm staying out of the 2" driver thing. Geddes teaches that you can't get there with that large a throat diameter. I'm pretty sure we CAN get there with 1.5", once we acknowledge that the phase plug is part of the deal.

Imaging and spaciousness are mutually exclusive only until we recognize the cues and learn to manage them. See Toole.... :yes:

Mike Caldwell

12-24-2009, 07:29 AM

I remember reading somewhere that a 2 inch throat causes phase cancellations at the extreme high frequencies. Also the four inch voice coil/diaphragm has it's high frequency physical limitations. The smooth 2445 diaphragm in tested showed that it physically distorted at high frequencies. The 2446, 2447, 2450, 2451 all have diaphragms with ridges in them to make them stiffer. The new JBL neo drivers in the Vertec cabinets are 1.5 throats and three inch voice coil/diaphragms.

mikeharris

12-24-2009, 08:19 AM

We havent tried these in 2445's...but when faced with the same problem in TAD4001's...we have substituted a Radian diaphragm in control room applications.

http://www.radianaudio.com/index.php?option=com_virtuemart&page=shop.browse&category_id=32&Itemid=40

The old school way was a 2420 with no horn

spkrman57

12-24-2009, 08:52 AM

Good things happen when you can find a system, understand what it will or will not do for you, and when you realize the "best bang for the buck" solution and can accept it, then you can start to enjoy the music.

I happen to listen to music solo, so a sweet spot on the couch is measured out and tested by listening with my ears to know where it is. I use Edgar horns and happen to not mind as much as others having to sit still in a specific postion. The Edgars beam heavily with the 650hz being the worst, and the 350hz being the least picky with the sweet spot location.

So I know what widget is saying with the "head in the vise" statement. But I will say that with horns that have different dispersion characteristics that I can't get the imaging in 3D that the beaming Edgars present to the listener, similiar to the headphone listening experience.

I also am willing to use a Altec 299 1 - 3/8" drivers which limp up to 18khz, but with EQ performs well enough to me to not require a tweeter. Of course I'm over 50 yrs old and my ears have been abused with factory noises/computer room noises/living on board US Navy ship and near military airstrips. Not to mention the bars with loud amps/spkrs that were cranked out in the 70's and 80's.

All I can say is to come up with a reasonable goal with your audio system. Be realistic in what it can do in your house and decide if that will work for you.

I'm a audio fanatic in that I (used to) own lots of "pieces-parts" to tinker with for my learning enjoyment. Sometimes it is harder to put a system together when you have too many parts to try out!!!

Ron sends....

Yep... those Edgar horns do image remarkably. I'd hate to give that up, but the head in the vice effect is pretty bad.

Good luck finding just the right compromise. There is no way you will find a perfect two-way with excellent imaging, but with luck you may get acceptably close.

Widget

Mr. Widget

12-24-2009, 09:50 AM

So I know what widget is saying with the "head in the vise" statement. But I will say that with horns that have different dispersion characteristics that I can't get the imaging in 3D that the beaming Edgars present to the listener, similiar to the headphone listening experience.I wouldn't describe it quite in that way... I agree that the Edgars as well as a few other designs will produce an image that is almost holographic, but unlike headphones which place the "image" in the center of your head, the Edgars will place the "image" in front of you.

I will say this, I have heard the Edgars with JBL 375 drivers on them and 2405 slots above and the imaging wasn't harmed at all and the sound was quite spectacular... as long as your head was in the sweet spot.

Widget

Allanvh5150

12-24-2009, 11:27 AM

I will say this, I have heard the Edgars with JBL 375 drivers on them and 2405 slots above and the imaging wasn't harmed at all and the sound was quite spectacular... as long as your head was in the sweet spot.

Widget

Is there any other place to be?

cosmos

12-24-2009, 01:20 PM

Alright, I have heard from the jury and I believe, based upon its learned experience, that a two way simply won't give me good UHF response with a 2" driver.

Frankly, with all the years of music and industrial exposure, I doubt I can hear anything above 14KHz.

I own 2405s and 2404s

I guess I will use the 2445 with 2404 or 2405 on top.

Then I think I will attack the goal of a good two way with a 1.5" driver.

Now I need good 1.5" drivers and good 1.5" waveguides. Any suggestions?

Zilch

12-24-2009, 01:34 PM

Now I need good 1.5" drivers and good 1.5" waveguides. Any suggestions?PT waveguides and 2452H-SL drivers.

243x drivers can work, too; plenty of threads on doing that here.

Cheap (by comparison) and easy peasy? BMS 4555 drivers.... :yes:

Mr. Widget

12-24-2009, 04:11 PM

I own 2405s and 2404s

I guess I will use the 2445 with 2404 or 2405 on top.Why not... you can always try them with and without. Then you can decide for yourself. I personally prefer the 2405 by a wide margin, but you do need a pair that are in excellent shape. Many are not.

PT waveguides and 2452H-SL drivers.

243x drivers can work, too; plenty of threads on doing that here.

Cheap (by comparison) and easy peasy? BMS 4555 drivers.... :yes:Better still with a proper tweeter... for those who can still hear the higher frequencies. ;)

Widget

subwoof

12-25-2009, 05:02 PM

I built this HF/UHF combo for my rental house PA - it works quite well for it's app and the horn ( 2332 1.5" with the 2450SL ) gives all the volume and realism needed. the 2404 was added to eliminate the vertical dispersion issues the horn had above 10K.

I tried this with the square PT horn and 2431H driver but the extension wasn't there unless on axis. Perhaps this summer this can be resolved with proper filters. Currently they are 3 way active ( M533 ) with a 3105 clone.

The tent sale had these older rectangular horns for 30 bucks each sometime ago - however the newer PT horn ( 95H ) is the same size and only about 100 bux each.

And yes - some of older types still like to annoy bats. Use a tweet.

cosmos

12-27-2009, 02:26 AM

I built this HF/UHF combo for my rental house PA - it works quite well for it's app and the horn ( 2332 1.5" with the 2450SL ) gives all the volume and realism needed. the 2404 was added to eliminate the vertical dispersion issues the horn had above 10K.

You are one heck of a landlord!

I think I'll try the 2404s and I will also likely try a pair of Heils as UHF..

Thanks!

jf65

01-04-2010, 06:31 AM

Extending the 2445's response to ~20kHz is possible, but at the cost of about 10-12dB sensitivity.

I have made a passive equalizing network for the 2445/2360 combination with intended crossover at 600 Hz and response from 800 Hz to 20kHz within +/- 2.5dB. To achieve this, the 1kHz level must be damped about 12 dB.
The actual levels and crossover values may differ, but in principle the same is achieveable in other horns: like the 2380.

Hello, could you tell us more about this passive equalizing network, i have no idea about to do that.
thanks
Jean

jerv

01-04-2010, 01:36 PM

Hello, could you tell us more about this passive equalizing network, i have no idea about to do that.
thanks
Jean

http://www.audioheritage.org/vbulletin/showpost.php?p=271106&postcount=7

The network is shown near the bottom of the post.

Please bear in mind that even if this network may show that it is indeed possible to get to 20 kHz with the 2445 driver - this may not (as many other posters and myself point out) be the best overall solution.

The 2445 tend to sound a bit harsh used this way. And because of the 2-inch throat, the horn will beam at high frequencies. IMO, you're best off with a supertweeter.

jf65

01-04-2010, 02:58 PM

http://www.audioheritage.org/vbulletin/showpost.php?p=271106&postcount=7

The network is shown near the bottom of the post.

Please bear in mind that even if this network may show that it is indeed possible to get to 20 kHz with the 2445 driver - this may not (as many other posters and myself point out) be the best overall solution.

The 2445 tend to sound a bit harsh used this way. And because of the 2-inch throat, the horn will beam at high frequencies. IMO, you're best off with a supertweeter.

Thanks, i agree with you, i intend to use it up to 9000 HZ. My goal is to have a flat response from 600 to 9000. Now it's flat from 1000 to 3500 (Iwata horn) but goes down slowly after that limit and a tweeter (T 925) takes it from 6000 (it would be more at ease at 9000).
Tell me if i'm wrong, i think that the 1041 components are for compensation and the other ones the high pass crossover?

jerv

01-05-2010, 12:45 AM

...Tell me if i'm wrong, i think that the 1041 components are for compensation and the other ones the high pass crossover?

Mostly right. Though all components react with each others (and with the inductance in the driver's voice coil), the 1041 series LCR makes a shallow notch filter (Q=1.2), centered at about 5200 Hz. The C1011 and the parallell LR 1021 takes care of high pass crossover and dampens down the 800Hz-2kHz range. Don't use that crossover if you intend to use a supertweeter. It attenuates the 2445 much more than what is needed with a tweeter.

What you need to do, is to attenuate the 800-3500 Hz range. This can be accomplished with the combination of a notch and a shallow high-pass filter. This is my passive crossover for the 2445J on the 2360 horn, intended for crossovers at about 600 and 8000 Hz. The 2360 is a constant-directivity horn, but the Iwata (as far as I know) isn't. The 2445 on the Iwata should (in theory, at least) need less compensation than on the 2360. My guess is that the same filter topolygy can be used on the Iwatha horn, but components values may vary.

jf65

01-08-2010, 11:36 AM

Mostly right. Though all components react with each others (and with the inductance in the driver's voice coil), the 1041 series LCR makes a shallow notch filter (Q=1.2), centered at about 5200 Hz. The C1011 and the parallell LR 1021 takes care of high pass crossover and dampens down the 800Hz-2kHz range. Don't use that crossover if you intend to use a supertweeter. It attenuates the 2445 much more than what is needed with a tweeter.

What you need to do, is to attenuate the 800-3500 Hz range. This can be accomplished with the combination of a notch and a shallow high-pass filter. This is my passive crossover for the 2445J on the 2360 horn, intended for crossovers at about 600 and 8000 Hz. The 2360 is a constant-directivity horn, but the Iwata (as far as I know) isn't. The 2445 on the Iwata should (in theory, at least) need less compensation than on the 2360. My guess is that the same filter topolygy can be used on the Iwatha horn, but components values may vary.

Hello, i just know the basics of crossovers, so i don't understand how a high pass filter at 600 hz can dampen the 800-3500 Hz range. Besides, i noticed that your 2445 is a 16 Ohms version, mine is 8 Ohms, so i will have to work on components values and as you say the Iwata horn is also different; but i will follow the principle of your crossover.
best regards, Jean

jerv

01-08-2010, 03:50 PM

You are right: it cannot. But togehter with a notch, it can.

In this filter, the 1041 notch is centered at about 970 Hz with the given component values. Togehter with the low-Q high-pass filter (C1011 and LR 1021), the driver's response is about 6dB down at 600 Hz, initially with a very shallow rolloff so the 1000-1500 Hz level is about 2-3 dB down as well.

Sorry: the 8-ohm version will need other component values. Since I don't have frequency and impedance measurement files for the 8 ohm, I cannot simulate even approximate filter component values. The given values are for the 2445J (on the 2360).