4344 Active Project

[ayaboh]

Hi Guys.

I have been working on a 4344 project for a while, and done some real engineering and 3D modeling for this project. The idea is to make a fully active speaker system based on Hypex UcD modules (http://www.hypex.nl (http://www.hypex.nl/)), but be able to start out with a passive network and add amps and stuff later. The cabs have enough internal space to put all the network components, amps, powers, and transformers inside. The compartment for the electronics is part of the internal sub-structure, and can be accessed trough a door in the back panel.

I have constructed a sub-structure as a back-bone for the cabinets, and used a lot of bracings for the baffle and back plate. This should make the cabinet much stiffness and more stable than the original JBL.

The cabinets are deeper than the original 4344s due to the volume occupied by the amp/filter box, sub-struc and bracing, but the net volume is probably close to the original.

Some preliminary data:

Dimensions = 1050 x 635 x 410 mm (hwd)
Sub assy weight = 12,5 kg (fiber board with density 0,685 g/cm^3)
Braces weight = 3,25 kg (pine wood with approx density 0,25 g/cm^3)
Driver weight total = 23,5kg
Total weight = 81,55 kg

Total internal volume = 210,6 L
Dog box volume = 15,4 L
Filter box volume = 13,8 L
Sub assembly volume = 18,26 L
Bracing volume = 9,23 L
Net internal volume =153,91 L
Volume occupied by drivers and horn (except 2123) = 7,5 L (+-0,1 L)
Net internal volume with drivers in place = 146,41 L

Duct dia = 100 mm id

Parts used:
2235H, 2123H, 2420J with H-91 and Coral AL-603, 2405H

Take a look at the pictures and give me some feed-back on the design, construction, and baffle layout. The production drawings will be made later when I have necessary data for all the smaller parts.

I need some help with the duct length. I guess some of you guys out there know how to calculate this based on the given data.

You can find pics of the 3D-model in 1600x1200 here: http://www.designcut.com/4344

http://designcut.com/4344/lowres/4344-1.jpg
http://designcut.com/4344/lowres/4344%20Internal-2.jpg
http://designcut.com/4344/lowres/4344%20Internal-7.jpg

[johnaec]

'Looks very nice! Are the cabinet sides double-thickness? (Solid!). Are you allowing for some kind of amp ventilation?

Also - what program are you using for 3D modeling? ('Always interested...)

John

[Robh3606]

Hello ayaboh

Hey that look real nice!! I see you have the net volume figured. You can get WinIsd to try on your own.

http://www.linearteam.dk/default.aspx?pageid=winisd

The 2235 driver is preloaded. It is very easy to use and is shareware.

Will post duct length latter on but really where do you want to tune them?? Cabinet tuning will determine duct length. Do you want clasic flatest response or extended bass. I have my DIY 4344's cabinet tuned at 30hz, Attached is a quick look using WinIsd port lengths look to be 11.34 inches. Didn't check your internal depth to see it you have enough clearence. Figure you want about 3" or so behind them.

Rob:)

[Earl K]

- Beautiful Drawings !

- I'd redo the bracing in the woofer area.
- It's too close to the woofers magnet for my liking.
- From the looks of your pics , the bracing is blocking the magnets rear air vent. That's not a good situation .
- I'd like to see, 3" or 4" of clearance between that vent and any bracing immediately behind it .

:)

[matsj]

That looks really good .

I just got a question :). How did you get the speakers in the CAD program ? I got my Cad program yesterday ( solid works ) and i´m not so good at the program.

regards mats

[ayaboh]

'Looks very nice! Are the cabinet sides double-thickness? (Solid!). Are you allowing for some kind of amp ventilation?

Also - what program are you using for 3D modeling? ('Always interested...)
John

The sides are 25 mm plywood. They may look thicker because of the 45 deg end cuts. The UdC amp modules I plan to use are dgital and produce very little heat, and a small heat sink at the back door should be enough.

The modeling has been done in Autodesk Inventor.

[ayaboh]

Hello ayaboh

Will post duct length latter on but really where do you want to tune them?? Cabinet tuning will determine duct length. Do you want clasic flatest response or extended bass. I have my DIY 4344's cabinet tuned at 30hz, Attached is a quick look using WinIsd port lengths look to be 11.34 inches. Didn't check your internal depth to see it you have enough clearence. Figure you want about 3" or so behind them.

Rob:)

I want a flat response.

I have 16 inches of free space for the duct, so a length of 11.34 is a bit to much if it's necessary with 3 inches of space between the duct end and rear wall. Is it better to increase the diameter of the duct?

[ayaboh]

- Beautiful Drawings !

- I'd redo the bracing in the woofer area.
- It's too close to the woofers magnet for my liking.
- From the looks of your pics , the bracing is blocking the magnets rear air vent. That's not a good situation .
- I'd like to see, 3" or 4" of clearance between that vent and any bracing immediately behind it .

:)

Hi Earl.
Are you taking about the pole vent outlet, or is it that the bracing is to close to the magnet? I think the clearence is 2 inches, but I have to check my drawings to be sure.

[ayaboh]

That looks really good .

I just got a question :). How did you get the speakers in the CAD program ? I got my Cad program yesterday ( solid works ) and i´m not so good at the program.

regards mats

Hi Mats.

The models of the drivers are made from scratch.

[Robh3606]

I want a flat response.

I have 16 inches of free space for the duct, so a length of 11.34 is a bit to much if it's necessary with 3 inches of space between the duct end and rear wall. Is it better to increase the diameter of the duct?

Then you should download WinIsd. The tuning is off a couple of Hz so go higher to give you flat however you will loose a little extension down low. As far as the port you need to change the tuning, go to a single, or smaller ID to shorten the length on the ports. You can also angle them a bit with a partial elbow to keep the length as is.

Rob:)

[Mr. Widget]

The models of the drivers are made from scratch.Cool project! I like heavily braced designs too. Instead of 25mm plywood, I think you would find a slight improvement with 13mm MDF covered in 13mm birch ply. The bass will be slightly tighter with slightly less box loss.

If you don't mind, I'd love iges files of your drivers... I have been too lazy to build such nice examples.


Widget

[cvengr]

Nice job, ayaboh.

[matsj]

Is it possible to get a file on the 2235, it´s similar to my 2226 and the 2405 ?

Mr.Widget what drivers do you have on cad ?

Is here someone who have 2245,2446, 2450, 2382 on Cad files ?

regards mats

[4313B]

Please post an impedance curve of the finished system as well as a near-field response graph of the LF transducer. I'd recommend you build a single prototype and test it.

[ayaboh]

Please post an impedance curve of the finished system as well as a near-field response graph of the LF transducer. I'd recommend you build a single prototype and test it.

Sorry, I don't have the equipment for that. Is this necessary because of the internal bracing/sub struc or the increased volume? If this throws the impedance and response for the 2235 way off, it will be much easier to copy the original 4344 cabs instead.

[ayaboh]

Cool project! I like heavily braced designs too. Instead of 25mm plywood, I think you would find a slight improvement with 13mm MDF covered in 13mm birch ply. The bass will be slightly tighter with slightly less box loss.

If you don't mind, I'd love iges files of your drivers... I have been too lazy to build such nice examples.


Widget

Widget and Mats;

I can try to export the driver models to iges. I don't know if this will destroy the styles and formatting.

[Earl K]

Hi Earl.
Are you taking about the pole vent outlet,

-Yes, mostly that's my point .
- I'm concerned your bracing is too close to the vent outlet in the 2235s' magnet.
- JBL recommends at least 3 inches of clearance ( according to the 2226H spec sheet )



or is it that the bracing is to close to the magnet?
I think the clearence is 2 inches, but I have to check my drawings to be sure.

- Great drawings ! ( again )


:)

[4313B]

Sorry, I don't have the equipment for that. Is this necessary because of the internal bracing/sub struc or the increased volume? If this throws the impedance and response for the 2235 way off, it will be much easier to copy the original 4344 cabs instead.Yes, you're going to have to determine if that kind of bracing solution is going to negatively affect the response and tuning. Standing waves show up as blips in the impedance curve. When you tune an oscillator to those blips the resulting cacophony is usually quite irritating. You then try to damp them. I don't really have time to count how many standing waves this particular bracing solution is introducing right now but you probably get the idea. Usually the goal is to get the bracing to appear transparent to the volume or to get the standing waves so high in frequency relative to the intended badwidth that their effect is minimized.

[Mr. Widget]

Widget and Mats;

I can try to export the driver models to iges. I don't know if this will destroy the styles and formatting.I would expect your nice looking foilcals will be lost, but the basic geometry should be fine. I'd really appreciate it if you could give it a shot and send me a PM.




Mr.Widget what drivers do you have on cad ?
I have only done place holders... that look close from the front... I've never taken the time to accurately build a driver.

BTW: I use Solidworks too. I is great for machine parts, but kind of sucks as a surface tool and it hates complex fillets and the like.



Widget

[Mr. Widget]

Sorry, I don't have the equipment for that. Is this necessary because of the internal bracing/sub struc or the increased volume? If this throws the impedance and response for the 2235 way off, it will be much easier to copy the original 4344 cabs instead.Giskard makes a good point... all of those parallel surfaces can lead to some audible problems... every box has internal standing waves, it's the magnitude and frequency that are the issues.

One possible partial fix would be to angle each of your horizontal braces... it makes construction a bit more of a pain, but it does reduce the effect of standing waves.

As for measurement equipment, there is an inexpensive device that you can buy to measure your speakers' impedance. It is called the WT2.

http://www.woofertester.com/


Widget

[4313B]

every box has internal standing waves, it's the magnitude and frequency that are the issues.And to use this opportunity to answer the people who have posted that "some subwoofers are cubes". Invariably those cubes are quite small (meaning the standing wave frequency is higher as opposed to lower) and the low pass filters set so low, and are of sufficiently high order, that the standing waves are "adequately" attenuated. Think "signal to noise", where the standing waves are the noise.

[Ian Mackenzie]

It looks very nice.

One potential issue might be buzzes and other nodal resonances with all those compartments. The other thing is assuming you intend to add fibreglass to all surfaces, the additional surface area of all those internal panels will increase the volume of Fill and this may effect the final tuning and the apparent bass quality.

If you are at all concerned about noise, put the ports anywhere but the front baffle. Reflex ports can be a significant contributor of audible internal enclosure noise (via the vent)

[WTPRO]

Hi

Regarding the final tuning, another Woofer Tester feature is the ability to measure and analyze the in box driver impedance to come up with an effective box size and tuning. This example is pretty typical when using the Rem/Xem measurement and modeling. This driver in particular is representative of a driver with a high Rem/Xem. In this case I am comparing the simulators driver output to the measured driver output. The port in this case is somewhat isolated and on the back side of the box. The in air measurement was made using a Woofer Tester Pro (that I am madly trying to get finished for the Xmas shopping season, along with the Speaker Tester).

For those not familiar with Rem/Xem, you can think of 'Le' as being a variable where 'Le' is the effective inductance at 1 Khz. I highly recommend using the Rem/Xem data and simulator, but we also support Le and 'none' if you need to compare with older models or methods.

Another feature people may not be aware of is that the WT can also measure Cms variation with drive level. Please check our FAQ for further information on that topic.

It may also be worth mentioning that we are also now a distributor for the Harris Tech 'Bass Box' products.

Best regards,
Keith Larson, WTPro
Smith and Larson Audio
www.woofertester.com

[4313B]

Nice job Keith! :yes:

[ayaboh]

Widget and Mats (and other);

You can get the IGES files here http://www.designcut.com/4344/Iges.zip and driver foilcals here http://www.designcut.com/4344/Driver%20Foilcals.zip.

[ayaboh]

Yes, you're going to have to determine if that kind of bracing solution is going to negatively affect the response and tuning. Standing waves show up as blips in the impedance curve. When you tune an oscillator to those blips the resulting cacophony is usually quite irritating. You then try to damp them. I don't really have time to count how many standing waves this particular bracing solution is introducing right now but you probably get the idea. Usually the goal is to get the bracing to appear transparent to the volume or to get the standing waves so high in frequency relative to the intended badwidth that their effect is minimized.

Yes, this can be a problem, but have any of you looked inside a B&W matrix cabinet? They don't seem to be concerned about all these parallel surfaces. Another thing, (and I have not checked or tried to calculate this) is the space inside a cab like this, large enough to support standing waves at the frequencies we are talking about, or is structure borne waves the problem?

Anyway, I will finish a model of an original 4344 according to Ian's plans. It is nice to have a complete set of production drawings for this speaker, and as a bonus, all the errors and typos shows up when you make a full 3D model.

[ayaboh]

It looks very nice.

One potential issue might be buzzes and other nodal resonances with all those compartments. The other thing is assuming you intend to add fibreglass to all surfaces, the additional surface area of all those internal panels will increase the volume of Fill and this may effect the final tuning and the apparent bass quality.

I have been thinking about thin Neoprene foam mats on the dividers.



If you are at all concerned about noise, put the ports anywhere but the front baffle. Reflex ports can be a significant contributor of audible internal enclosure noise (via the vent)

All the original JBL monitors have the ducts on the front.

[Robh3606]

All the original JBL monitors have the ducts on the front.

That is because of how they would typically be mounted in use. You don't have those constraints. Unless you are going to actually make exact copies which you obviously are not with the bracing and active crossovers keep an open mind. There are better ways to do them including changing the layout of the drivers, particulary the 2235, and getting the vents off the baffle board and away from the 2122 as an example compared to the original layout. Ports on the back is a good option.

Rob:)

[matsj]

ayaboh thanks alot. I will take a look at it when i come home.

regards mats

[Mr. Widget]

Widget and Mats (and other);

You can get the IGES files here http://www.designcut.com/4344/Iges.zip and driver foilcals here http://www.designcut.com/4344/Driver%20Foilcals.zip.
Thanks! I just downloaded them, but am on my Mac... I'll have to wait until I boot up my PC to check them out... I'll let you know how it works out.


Widget

[Mr. Widget]

I don't have a rendering program, so I don't have surface mapping and other niceties... but your model came through beautifully.
Thanks a million... these will be a welcome addition!


Widget

[ABJ]

Hi ayaboh.

Are you cracy ?

Never have i seen such job for a hobby project. Really nice. I have downloaded your .ige/.xlo files for later use. I am beginner in inventor, and yes, i do use some of the same components.
My system is 077, LE104, 2123H, 2243H

Comments for your design :
Its allways a good idea to build a test cabinet, without all the nice details (and expencive), as Gisgard says. To hear if the electonic cabinet etc. will have some influence on the sound. As you might know, speakers and cabinets is a funny thing to work with. The smallest thing can destroy everthing.
You dont mention if you gonna put a amp. inside the boxes. If you intend to, re-consider. The vibration form the LF and 2123 can mess up your eletronics in time. (Solderings, components etc) but if its only the crossover, its a nice idea...or not. The chamber for electronic will...sound...what sound i dont know, but it will "sound" in one way or another. But i agree, nice thing to have such door for crossover

Is it possible to get the "what ever" .xxx files for the Inventor for me to "play" with ? I have used ACAD 2000 for manny years, and now ill like to "move on" if you know what i mean..

nice job...

Regards, Anders

[matsj]

It was no problem to fit them into solid works, many thanks again.

regards mats

[ayaboh]

Hi ayaboh.

Are you cracy ?

Never have i seen such job for a hobby project. Really nice. I have downloaded your .ige/.xlo files for later use. I am beginner in inventor, and yes, i do use some of the same components.
My system is 077, LE104, 2123H, 2243H

Comments for your design :
Its allways a good idea to build a test cabinet, without all the nice details (and expencive), as Gisgard says. To hear if the electonic cabinet etc. will have some influence on the sound. As you might know, speakers and cabinets is a funny thing to work with. The smallest thing can destroy everthing.
You dont mention if you gonna put a amp. inside the boxes. If you intend to, re-consider. The vibration form the LF and 2123 can mess up your eletronics in time. (Solderings, components etc) but if its only the crossover, its a nice idea...or not. The chamber for electronic will...sound...what sound i dont know, but it will "sound" in one way or another. But i agree, nice thing to have such door for crossover

Is it possible to get the "what ever" .xxx files for the Inventor for me to "play" with ? I have used ACAD 2000 for manny years, and now ill like to "move on" if you know what i mean..

nice job...

Regards, Anders

Hi.

It's no problem to make the model files available, but you need Inventor ver 10.0 or later to use them (and a high-speed connection). I have been re-thinking the whole project, and I'm currently working on a model of an "original" 4344 based on Ian's (and others) drawings and documentation. I don't have the necessary equipment to go all the way with this project, so I think this is the best solution for me right now. As you know, an original 4344 is not exactly a flimsy project either.

[porschedpm]

Just to be a troublemaker and stir up the pot. But if you're going to be constructing the cabinets from the ground up, have you considered using the 18" 2245H low frequency driver instead of the 2235H and build a 4345 instead of the 4344. All other drivers are the same for both. The 4345 is to be considered to be the better sounding of the two.

[hjames]

Just to be a troublemaker and stir up the pot. But if you're going to be constructing the cabinets from the ground up, have you considered using the 18" 2245H low frequency driver instead of the 2235H and build a 4345 instead of the 4344. All other drivers are the same for both. The 4345 is to be considered to be the better sounding of the two.

Consider the 4345 has a 9 cubic foot cabinet = 40 x x30 x 18 1/2 (roughly)
while the 4344 uses a 5.5 cubic foot cabinet = 41 3/8 x 25 x 17 1/8 inch


All that extra wood makes it a bit heavier, too!
JBL 4345 - 229lbs vs 185lbs for JBL 4344

But that said, I hear there is nothing like the low end of a 4345 -

[ayaboh]

Just to be a troublemaker and stir up the pot. But if you're going to be constructing the cabinets from the ground up, have you considered using the 18" 2245H low frequency driver instead of the 2235H and build a 4345 instead of the 4344. All other drivers are the same for both. The 4345 is to be considered to be the better sounding of the two.

I already have the 2235s, so it has to be a 4344 this time. Later, - who knows? The production drawings for this project is more or less finished, and will be ready for download in a few days for those of you who are interested.

[Ian Mackenzie]

Better to Err on the side of caution with the WAF !!

Not to mention the nagging..... nag ..Nag ...nag...NAG.:banghead:

[Ducatista47]

But that said, I hear there is nothing like the low end of a 4345 -

The balance is difficult to top, too. A real all around pleasing speaker to live with. They can sound very intimate as well as being able to produce a "Big" sound. The integration of drivers is so good, you can't tell that there is more than one transducer at work. Surprisingly, they don't require a large room to do their work. After all, they are monitors.

Clark

[ayaboh]

Has anyone here actually heard the difference between these speakers, - I mean at the same time, in the same room, and with the same electronics? The only differences between them are the LF drivers and cabinet size, and possibly the network. I guess it's possible to get a bit extra extension and slam from a 2245, but how much does this mean in a normal living room?

[Ian Mackenzie]

I have not heard them at the same time but I have previously built a 4343 system with a 2235H some 26 years ago and currently run a 4345 clone.. that last few years. There are differences and this has a lot to do with the way the woofers couple to the room but the 4345 conveys a bigger sound.

There are no two ways about that.

I sometimes ran a second 2235H bin as a slave with the 4343 but that was only under difficult circumstances.

The 4344 is a more domestic system in terms of size and should be more than adequate for most home situations. Make sure you use a suitably proportioned amp on the 2235H and it will be fine (ie 250-600 watts). The LF response of the 4344 is very similar to the 4345.

[4313B]

Has anyone here actually heard the difference between these speakers, - I mean at the same time, in the same room, and with the same electronics? The only differences between them are the LF drivers and cabinet size, and possibly the network. I guess it's possible to get a bit extra extension and slam from a 2245, but how much does this mean in a normal living room?Unless you have the option to hear both you're not going to know that you're missing anything by going with the 2235H. I've used the LE14H-1, 2235H, and 2245H extensively since their debuts and I think they are all great LF transducers.

You can't get a single 2235H to outperform a single 2245H. There really is no substitute for cubic inches. The name of the game is to move as much air as possible with minimum excursion (don't get me started on these retarded, massive xMax transducers). At low volumes I think the 2245H offers the best resolution. The LE14H-1 is a close second followed by the 2235H.

Space is space and not everyone has the room for two 10 cu ft boxes in their listening room. If I were to do it all over again I would probably go with the LE14H-1 due to the smaller physical volume requirements and the fact that, over the years, I've actually come to like it better than the 2235H. Yes, I fully realize that an LE14H-1 or 2235H is not going to give me the same performance as a 2245H. Oh well. I'm pretty certain that I'd somehow make it through and be able to enjoy what I came up with. Worst case, the fireplace is in the same room.

[Ducatista47]

The name of the game is to move as much air as possible with minimum excursion (don't get me started on these retarded, massive xMax transducers)

OK, I'm sure someone besides Giskard can fill me in. I joined this site too late to see the flowering of the Sub1500. Was it a car speaker or what? And while this does require an opinion, and this is what I could not find by searching, what is it good for and what is it not good for? Is it or is it not hi-fi? I do know one of our members is using it for a very critical application.

Sorry for the newbie/latecomer question. I have two 2245H baskets that need recones and am curious what way to go for which possible use. Does the Sub1500 not require a much smaller box? I don't have anything specific in mind yet, so I seek generalized answers.

BTW, Giskard (Very, very nice to have you back, please stick around :)), most experienced, intelligent listeners I know think the LE14 series are the best LF speakers for home hi-fi use they have ever heard, while different a tie with my 2245H's in quality of sound. So yes, they must be close. But I have to agree the larger single cone would resolve better than two (smaller) discrete sources, be they both active or active/passive radiator.

But then there is the DD66000! Two woofers which do have some signal overlap, electrolytic and Solen caps and it sounds better than anything else. :blink:



Thanks, Clark in Peoria

[4313B]

OK, I'm sure someone besides Giskard can fill me in. I joined this site too late to see the flowering of the Sub1500. Was it a car speaker or what?It's the REVEL 15 transducer originally used in the REVEL GEM sub. The designer runs two of them at home. It was then given a "JBL number" and used in a few overseas subs, a ported version in Japan and a sealed version in Europe. It's replacement is the W1500H used in a new sub for Japan as well as the 1500 Array.

[johnaec]

I joined this site too late to see the flowering of the Sub1500. Was it a car speaker or what?By "retarded xMax transducers" I believe Giskard was referring to the WxxGTi and similar series of car subs, (pro series of DD LF speakers also?). They're capable of huge xmax values, but are primarily designed for a limited range of high-volume reproduction at low frequencies, and not necessarily the highest fidelity, (pro series also?). Some people report good results using them for subs in home, but I don't think any serious testing has been done with this, other than frequency response.

The Sub1500 was specifically designed for high fidelity subwoofer use, but its transient response, (and phase?), are likely not equal to other drivers with less mass. And while it also has a high xmax, I doubt Giskard included this driver as a retard, (though I've been wrong before...).

John

[4313B]

By "retarded xMax transducers" I believe Giskard was referring to the WxxGTi and similar series of car subs, (pro series of DD LF speakers also?). No, Jerry built the WxxGti Series to kick ass and eat the piece of paper the names are written on in competition systems. They hold their own in that arena. I was actually thinking specifically about this while I was typing about retarded xmax transducers :p -

http://www.carstereo1.com/index_broadband.html

I'm running a pair of SUB1500's right now along with the 4355 12-inch 3-ways.


But then there is the D66000!I'm waiting for someone to put the 4435 bottom end with a 4345 10-inch 3-way or 4355 12-inch 3-way.

[Guido]

I'm waiting for someone to put the 4435 bottom end with a 4345 10-inch 3-way or 4355 12-inch 3-way.

Please help with the 4345 Be (2435) tests from this other thread and I'll jump on this.
The 4435 bottom end is already here :yes:

[4313B]

subwoof will send me the 1.5" exponentials as soon as he has time.

[mikebake]

I'm waiting for someone to put the 4435 bottom end with a 4345 10-inch 3-way or 4355 12-inch 3-way.
Yer killing me: that's great!:applaud:

[Robh3606]

Please help with the 4345 Be (2435) tests from this other thread and I'll jump on this.
The 4435 bottom end is already here :yes:

Guido I thought you were running 2435's on the PHxxx?? horns in your 4435's???

Rob:)

[Guido]

Guido I thought you were running 2435's on the PHxxx?? horns in your 4435's???

Rob:)

Nope!
At present I'm using 2344/2426 with D8R275ND Aquaplas dias. No network mods necessary.
GREAT!

I haven't had the time to finish the networks for the PH316/2435 combo.

At present I'm working with a horn speacialist to modify the PH316 throat equal to the 2344 throat to have better dispersion and end up with a quasi 1,5" 2344.