4430 + 2123H : a way to improvement.

[screw_squirrel]

Good morning everybody,

I have bought a pair of 4430 2 years ago and played it with a single McIntosh MC2255 amplifier. One year after I upgraded the system by bi amplifying it with JBL6260 and MC2255. The result is like day and night, the system is much better. But… I have read that I could improve it by adding a medium speaker with no modifications on the basic speakers.
So I have found a pair of 2123H and I want to build the enclosure and buy the right crossover to achieve this project.
Knowing that I think use the 2123H from around 300Hz to 1000Hz, to make a good blend with the couple 2425H + 2344A

My questions are:

- What dimensions and volume for the enclosure with 2123H?
- Should the front side of the enclosure be like à square 30cmx30cm or can it be 55cm wide to fit the 4430 width dimension?
- Is the M553 crossover a good choice?
- Should I prefer a digital crossover with multi frequencies and slopes choice like dbx PAdriverack.?
- Can I run two active filter 5235 in serial with 300Hz and 1000Hz cards?

A lot of questions as you can see…
Can someone provide good advices to about it?
Thanks to anyone
Eric

[Robh3606]

Hello Eric

I ran an E-145, 2123, 2425 2344 combo using an M553 for a while. It was quite nice soundind. I now use a PTH1010 and a 2435 for the top end.


What dimensions and volume for the enclosure with 2123H?

I think it's .25 cubic ft but check the data sheet and Drew Daniels system on the site.


Should the front side of the enclosure be like à square 30cmx30cm or can it be 55cm wide to fit the 4430 width dimension?


I would match the 4430 width as far as the baffles concerned.


Is the M553 crossover a good choice?


Works for me.


Should I prefer a digital crossover with multi frequencies and slopes choice like dbx PAdriverack.?

Yes as long as you are going to actually use it.


Can I run two active filter 5235 in serial with 300Hz and 1000Hz cards?


I think you are better off with the steeper 24db slopes


I think use the 2123H from around 300Hz to 1000Hz

I would go higher on the 2123 so you match the directiviy of the 2344 at crossover.

Have Fun

Rob:)

[pos]

- Can I run two active filter 5235 in serial with 300Hz and 1000Hz cards?
You need to cross the 2123/2344 at 1600Hz to match directivity:
http://audioheritage.org/vbulletin/showthread.php?t=24621#post247137

Concerning the 2235/2123 crossover point, that is much more room and placement dependant, so your are up for some experiments:
http://www.audioheritage.org/vbulletin/showthread.php?p=276134#post276134

In both cases I would use Linkwitz-Riley slopes (12 or 24db/oct) :
http://www.rane.com/pdf/ranenotes/Linkwitz%20Riley%20Crossovers%20Primer.pdf

the good thing is that at 1600Hz you do not need to take the horn rolloff into account, so the electrical slope will more or less match the acoustical one.

[screw_squirrel]

Thank you Rob and Pos for your good advices! ;)
Here's a picture of the first version of my project: 14 L box (1/2 ft²).
I did it this morning at work during lunch time. 1/4 ft² seems to be a bit small. I'll post another version with width fiting 4430 size and a place for a future 4 ways extention with a 2405H or 2404H...
Conserning the active crossover I don't know what to think... My expérience is too small, and I know quite nothing about electronics.I'll try to find a M553 crossover, even if I think a driverack PA could help me better because it's more flexible.
Eric

[1audiohack]

You are headed down one of my favorite paths.

I have liked my 2123's enclosures best between 0.5 to 0.75 cubic feet loosely filled with fiberglass.

I agree with the crossover point 2123 to the 2344, 1300 to 1600 Hz. As a bonus, the 2123 and the 2344 (at least with a 2425 or 2426 driver) are right in time with each other at those frequencies if mounted on the same vertical plane so no delay or displacement is required. That makes use of a M553 just fine. Get them together as close as possible, its as easy a combination to make sound great as any I have used.

[screw_squirrel]

Well, I have simulated with WinISD the 14 liters box and the 21 liters box for 2123H with Highpass and Lowpass, Linkwitz-Riley (LR-4) crossover, as if it was done by M553 at 300Hz and 1600Hz. And the result is nearly the same... :blink: Something must be wrong in here but I don't know what exactly. I think I have missed one parameter somewhere...:banghead:



http://nsm03.casimages.com/img/2010/03/06/100306102304607125572876.gif

[pos]

14L is already quite big for the 2123H (Qtc ~0.5), so going bigger does change much (especially with a 300Hz cutoff)
You should also consider the volume taken by the driver, so a 14L box will result in a 11L net volume.

For the box layout I would try to keep the 2123H vertically aligned with the 2344 (the 2404 position does not matter much).
I you use symetrical L-R filters it would also make sense to try to align the 2123 and 2425 (the 2123H/2235H alignement does not matter nearly as much) diaphragms, but it will be impossible (or counterproductive) with such a shallow driver. I you cannot time align them geometrically maybe you should use a digital crossover...
You can also try different slopes in WinISD Pro and look at the delay at the crossover point (but keeping the crossing point at -6dB for both drivers). For example a 15cm compensation is equivalent to a 0.44ms delay. That means that if the 2123H and 2425H diaphgrams are 15cm appart (that would need to be mesured...) you should try to achieve a -6dB crossing point with the 2123H having 0.44ms more group delay than the 2425H.

[ratitifb]

increase the internal volume of the dogbox from 14l to 21l is negligible compared to the T/S parameters of the 2123 and its own free frequency response ...

[Robh3606]

even if I think a driverack PA could help me better because it's more flexible.

If that is what you are more comfortable with go for it. The idea with using the 2123 with the 2344 is to get them as close as you can. You are also going to be in a seperate box you can move back and forth so "time alignment" is something you can adjust easilly. I had mine both mounted to the same baffle plain and it sounded fine.

Rob:)

[screw_squirrel]

Thanks Pos and Ratitifb for your help and support. This shows the initial project with an internal 14 Liters box (net), without speaker. 2123H and 2344A are verticaly aligned and in the depth also. My concern now is to find a M553... Or pick a 4 ways crossover to be ready for the next step => 2404H tweeter


http://nsm03.casimages.com/img/2010/03/06/100306030442607125574501.jpg

[4313B]

JBL used these drivers in "oversized" boxes to get Fc as low as practical since it plays hell on any passive filter components. If you're going active then the box can be made smaller than the usual 0.5 cubic feet.

[screw_squirrel]

If that is what you are more comfortable with go for it. The idea with using the 2123 with the 2344 is to get them as close as you can. You are also going to be in a seperate box you can move back and forth so "time alignment" is something you can adjust easilly. I had mine both mounted to the same baffle plain and it sounded fine.

Rob:)
Nice alignment Rob! :applaud: On my project I have 330mm between 2123H axis and 2344A throat. I wish I could do better. :D

[screw_squirrel]

JBL used these drivers in "oversized" boxes to get Fc as low as practical since it plays hell on any passive filter components. If you're going active then the box can be made smaller than the usual 0.5 cubic feet.
Thank you for this information. ;)

[screw_squirrel]

Good morning.
As it seems not to be easy to find a M553 crossover at decent price, I am looking further for a backup solution involving my two JBL-UREI 2 ways active crossovers (5234A and 5235) in serial (cascade). That means reproduce 4 original blank cards with 18dB slopes based on 51-5130, and fill it with the correct componants according to 5235 manual. But... I have a big probleme I can't find the reference for the 6 pins connector .
If someone has the answer please help me! :D
Attached a new diagram with butterworth 18dB slopes @ 300Hz and 1600Hz.
Eric

http://nsm03.casimages.com/img/2010/03/07/100307125130607125580484.gif

[pos]

I would stick with L-R filters, at least for the 1600Hz crossover.
Maybe you should get a digital crossover, try different slopes and time delays, and then decide what to do.
I see you are in Paris, I can borrow you a DCX2496 to try if you want (I have a pair).

[ratitifb]

But... I have a big probleme I can't find the reference for the 6 pins connector .
If someone has the answer please help me! :D:p

Molex 0.156" PCB Connector, 6 pins.

Mouser #538-09-52-3063

at www.mouser.com (http://www.mouser.com/)

[screw_squirrel]

I would stick with L-R filters, at least for the 1600Hz crossover.
Maybe you should get a digital crossover, try different slopes and time delays, and then decide what to do.
I see you are in Paris, I can borrow you a DCX2496 to try if you want (I have a pair).
Thank you for your proposal Pos ;)
But First of all I have to build the enclosures and the speakers are still in the US where my sister lives.

[screw_squirrel]

:p

Molex 0.156" PCB Connector, 6 pins.

Mouser #538-09-52-3063

at www.mouser.com (http://www.mouser.com/)
Thank you Ratitifb. I have found out that it was a Molex part this afternoon but didn't know where to buy it!

[RichK]

Status on the project? I'm all ears!


Rich

[screw_squirrel]

Status on the project? I'm all ears!
Rich
Well, I have received the 2123H from the US. I didn't thought it was so heavy. I have also bought a M553, I think it will be easier to find the correct balance between the speakers with this item. Now I've got to find the right guy to built the enclosures. For this I am still wondering wich volume is the good one because I have many answers about it. The JBL literature says between 7 and 14 liters (0.25 to 0.5 cubic foot) but a gentleman from Belgium who optimise JBL items said to me 21 liters sounds better... So i am a little bit confused about this. If someone could help me to find out the right volume I would be happy.
Eric
New image with possible extension with 2404H.
http://nsm03.casimages.com/img/2010/03/22/100322054105607125682398.jpg

[4313B]

The JBL literature says between 7 and 14 liters (0.25 to 0.5 cubic foot) but a gentleman from Belgium who optimise JBL items said to me 21 liters sounds better... So i am a little bit confused about this. If someone could help me to find out the right volume I would be happy.Isn't that your job as the system designer? :blink:

I do recall that 14 liters is already "oversized" for that driver, anything bigger simply moves Fc down even further which you probably wouldn't care about if you are going active instead of passive.

[screw_squirrel]

Isn't that your job as the system designer? :blink:

I do recall that 14 liters is already "oversized" for that driver, anything bigger simply moves Fc down even further which you probably wouldn't care about if you are going active instead of passive.
Yes it is... But I've read so much information those days that I don't know what to think. It's like a brain wash! Too much information kills information! Thanks to put me on the right tracks. 14 liters max! ;)

[4343]

Good morning.
As it seems not to be easy to find a M553 crossover at decent price, I am looking further for a backup solution involving my two JBL-UREI 2 ways active crossovers (5234A and 5235) in serial (cascade). That means reproduce 4 original blank cards with 18dB slopes based on 51-5130, and fill it with the correct components according to 5235 manual. But... I have a big problem I can't find the reference for the 6 pins connector .
If someone has the answer please help me! :D

I found the schematic here:

http://www.audioheritage.org/vbulletin/showthread.php?t=666&highlight=51-5130

and here with pics of the cards:

http://www.audioheritage.org/vbulletin/showthread.php?t=980&highlight=51-5130

HTH:D

[screw_squirrel]

I found the schematic here:

http://www.audioheritage.org/vbulletin/showthread.php?t=666&highlight=51-5130

and here with pics of the cards:

http://www.audioheritage.org/vbulletin/showthread.php?t=980&highlight=51-5130

HTH:D
Hi, I have changed my mind since I have found a M553! :D

Here's an updated view of my project :bouncy::
45314 (http://nsm03.casimages.com/img/2010/04/22/100422030930607125883894.jpg)

[screw_squirrel]

Some new pictures:
45879
45880

[screw_squirrel]

So the project is finished now and working well. The 2235H working only until 300Hz @24dB is a giant step to a better sound. I think this is one of the best improvement on the 4430. It brings more clarity to the bass.
Here comes the pictures, I hope it will be helpful.

Phase I: The built

http://nsm03.casimages.com/img/2010/08/01/100801074111607126500168.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074111607126500169.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074111607126500170.jpg
http://nsm03.casimages.com/img/2010/08/01/100801075945607126500399.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074112607126500171.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074112607126500172.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074112607126500173.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074112607126500174.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074112607126500175.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074112607126500176.jpg

[screw_squirrel]

All of this was made by a French carpenter from Brittany named Antoine Rouault (http://rouaultacoustic.wordpress.com/) very good job! :D

http://nsm03.casimages.com/img/2010/08/01/100801074113607126500177.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074113607126500178.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074113607126500179.jpg

[screw_squirrel]

Well If I had a larger room the sound will be even better. But this is the next step...

http://nsm03.casimages.com/img/2010/08/01/100801074113607126500183.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074114607126500186.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074114607126500187.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074347607126500223.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074348607126500224.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074348607126500226.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074349607126500230.jpg
http://nsm03.casimages.com/img/2010/08/01/100801074350607126500231.jpg

[Titanium Dome]

Hmm. :hmm: Your project makes me want to do something like that, but you actually did it! Looks nice. :yes:

[4313B]

So the project is finished now and working well.Very nice! :)

[screw_squirrel]

Hmm. :hmm: Your project makes me want to do something like that, but you actually did it! Looks nice. :yes:
Well go ahead! It's a great improvement to the 4430.
Basicly these speakers are very good. If you bi-amp them, they are going better. And with this trick it's wonderful. As I've say, I have cut at 300Hz and 1000Hz. I've try at 1400Hz but the tonality was not as good (Angus Young guitar wasn't sounding right). So I prefered to let the couple 2426H + 2344A sing along. If you want to go further you can use a four band divider and add a 2405H or a 2404H for the high frequencies. And you 'll get a super 4344.
Now I have to work on the amplifier for the bass. The BGW 750G is good but I am sure that I can find something much tighter and deeper.
If you want me to send you the drawings I have done, please ask by P.M.

Eric

[screw_squirrel]

Very nice! :)
Oh thank you! :D
It took quite a long time but it was worth. I think this improvement is much better than adding a 2405H as I 've seen before. Because the main axis of improvement on this very good speaker is reducing the amount of frequencies done by the 2235H. and the 2123H is one of the best between 300Hz and 1000Hz or above.

Eric

Edit: Maybe I could have used a 2" horn instead of the 2123H, but I live in a small flat and I think it would be too much!:rotfl:

[pos]

Very nice indeed!
Have you done some frequency response measurements?


I have cut at 300Hz and 1000Hz. I've try at 1400Hz but the tonality was not as good
The 2123H has a rising response, so you have to take it in account when designing the crossover, otherwise you end up with either a hole under ~500Hz, or a bump above ~1000Hz. This is not easy to accomplish with an analog active crossover: you have to be able to spread the filter poles at different frequencies (for example one pole at 900Hz and three poles at 1400Hz), as is done in the 4344mk2 network. Even a DCX2496 does not allow to do that: you have to resort to using parametric EQ...
The 2122H would be easier to use with a simple crossover, as its response is essentially flat between 300Hz and 1200Hz.

[Robh3606]

Those came out really nice! That's one of my favorite combinations a 2122/2123 with the 2344. They sure do sound good together.

Rob:)

[screw_squirrel]

Very nice indeed!
Have you done some frequency response measurements?
The 2123H has a rising response, so you have to take it in account when designing the crossover, otherwise you end up with either a hole under ~500Hz, or a bump above ~1000Hz. This is not easy to accomplish with an analog active crossover: you have to be able to spread the filter poles at different frequencies (for example one pole at 900Hz and three poles at 1400Hz), as is done in the 4344mk2 network. Even a DCX2496 does not allow to do that: you have to resort to using parametric EQ...
The 2122H would be easier to use with a simple crossover, as its response is essentially flat between 300Hz and 1200Hz.
Hi pos!
No not yet i've just miss this (http://cgi.ebay.fr/ws/eBayISAPI.dll?ViewItem&item=250660417918&ssPageName=STRK:MEWAX:IT) on eBay a couple of weeks ago...
But for the moment I've just ajusted the sound "by ear" with a friend of mine who is a sound ingeneer at Opéra Bastille Paris. ;)
The résult is not bad actualy, but could sure be better. My room is small, as you can see on pictures, but hasn't great defaults... The main problem remain the sound level. I can't listen too loud. But in an appartment who can...?
First I am willing to buy a ECM 8000 in order to mesure the sound produced by my system. My friend had a professional sound card with XLR entries and a RTA software. The second step would be to correct the sound with a BEHRINGER DEQ2496.

Eric