Building my dream system, need lots of advise!

[jmpsmash]

Hi all,

Before I start, let's start with an introduction and a little history. I have been dabbling with audio on and off for around 2 decades. My mother, of all ppl, is a music lover and an audiophile who got me hooked. My main music category is classical (symphonies/orchestral mostly) and a little jazz and no other genres.

Over the years going to shows and visiting friends (of friends) there are a couple of systems which really made an impression on me. One is a system with a 2 way, a pair of 2397/375 + 136A housed in a rather large translam cabinet. The bass definition was unreal while the texture coming from the Smith horns were that of a dream, these were driven by a 200W SET amp with a soccer size radio output tube. Another system is a multiway, with 3 horns and 2x JBL 15" and 2x JBL 18" housed in separate pro cabinets, on EACH SIDE. driven by DSP crossver and Crown amps. Playing orchestral music, the scale, dynamics, and at the same time, completely effortless presentation was phenomenal. I couldn't stop smiling for hours.

While the latter system is too large to be housed in my own home, my dream is to bring the essence of both systems and fuse it into my own dream system.

With a fair amount of reading and research, I think I have come up with something that is feasible and reasonable in my listening space (20'x20' room).

I have, as some might seen, started gathering pieces already. I have acquired a pair of 2397 and a pair of 2441. I also recently got a pair of stilll in the box NOS pair of 2123H (just in case I want to do a 4 way), a pair of Fostex T925A (though I am tempted by a pair of 2405 but not sure the condition). These will be driven by a minidsp crossover SPDIF unit that feeds separate DACs and a mix of class D amps (NC400 and Nuprime).

I have most of pieces already and what is missing is a good woofer and cabinets. I have currently a quad set of Pure Audio Project OB15Neo in open baffle leftover from a previous project but I find them really not up to par when it comes to clarity and definition. there is enough volume and speed but rather woolly in the presentation.

So in the end, I think I want to end up with:

Fostex T925A
2441+2397
2123H (maybe?)
2235H

crossover point is planned to be 5kHz, 1kHz, 250Hz

I am pretty new to this, and for the 2123H and 2235H I need to build the cabinets. From what I read so far, they need to be 0.5 cuft and 5 cuft respectively.

I have a few questions...

Is the plan a sound one? Am I heading the right direction to good sound?

What I am not sure is the 2123H, do I need it? Can I do away with it and cross at 5kHz and 950Hz? Is that stretching the 2441/2397 and the 2235H?

I still haven't found a 2235H, what is a reasonable price for it?

is putting a Fostex in a predominant JBL system heresy? I find it harder and harder to find good condition drivers and buying new might be less headache though a pair of 2405 popped up recently which looked like in good condition. Should I go for those instead?

I have only build open baffles speakers up till now. There will be quite a lot of learning ahead for the cabinets.

Anything else I need to know?

[Dr.db]

You are heading the right direction, for shure! ;)

The JBL 2441/2397 is a great midrange combination and the 2123 a stellar midbass/midrange woofer. I love both of these units!
Using a 10" midwoofer and building a 4-way or building a 3-way without it is an ongoing discussion on this forum. I haven´t solved this decision for myself yet. I will suspend my ideas on this topic and leave it up for the others....

The transition between the Fostex 925 and JBL 2397 horn will be a problem. Not so because it is a Fostex, rather because it´s dispersion characteristics won´t match these of the 2397 horn.
The Fostex is aprox. 60° if I´m not mistaken. The 2397 is 140°.... For a smooth transition these should be closer together. That is why I would prefer the JBL 2405 with 120° dispersion. Or the Fostex 825...

[Ian Mackenzie]

Your ideas are good.

The gap is the acquisition of the 2235 woofers.

Talk to Edgewound at Upland Loudspeaker Repairs about getting old of some woofers or building up some woofers.

If you go with the dual 2235H woofers try them initially with the Smith Horn.

The Smith horns 2 inch throat limits diffraction up to 10,000 hertz.

Cut the Fostex horn over from 8000-14000 with a 3rd order filter. GT recommends a 3rd order high pass filter .

The Smith horn may not need a low pass filter.

Don’t over think the design. Start off with the basics and evolve following extensive listening.

Get the mini dsp USB mic and download REW for your initial measurements.

The 2397 should be smooth enough without any additional EQ.

Use a 12 dB LR filter on the woofer and horn @800 hertz (acoustic slopes) and either apply phase compensation by shifting the horn crossover point or adding delay to the woofers.

Don’t stress about the beaming of the woofers in the mid range too to start with.

Once you get this going you can then add the 2123H 10 inch mid cone.

Crossover at 250- 900 hertz with 12 dB LR (acoustic) slopes and apply phase compensation or delay.

As l mentioned you will need to measure the drivers individually and the crossover summation in order to get in the ball park. With some work you should be able to get a response +-3 dB from 35-20,000 hertz.

There are some good enclosure construction references in the Library and the technical references forum.

Use a suitable protection capacitor on the horn and the Fostex driver.

Tune the woofers enclosure to 30-34 hertz in 5 cu ft3

Enjoy your project.

[jmpsmash]

You are heading the right direction, for shure! ;)

The JBL 2441/2397 is a great midrange combination and the 2123 a stellar midbass/midrange woofer. I love both of these units!
Using a 10" midwoofer and building a 4-way or building a 3-way without it is an ongoing discussion on this forum. I haven´t solved this decision for myself yet. I will suspend my ideas on this topic and leave it up for the others....


Glad to have some reaffirmation that those are good choices! for 3-way / 4-way, that needs to be investigated. If anything, more things to play with! My plan is to build separate cabinets for the 2123H. That, with addition to using active DSP crossover, will allow me to mix in the midrange and adjust the crossovers within seconds and do a proper comparison.



The transition between the Fostex 925 and JBL 2397 horn will be a problem. Not so because it is a Fostex, rather because it´s dispersion characteristics won´t match these of the 2397 horn.
The Fostex is aprox. 60° if I´m not mistaken. The 2397 is 140°.... For a smooth transition these should be closer together. That is why I would prefer the JBL 2405 with 120° dispersion. Or the Fostex 825...

What's the effect of having mismatched dispersion? is it for off axis, like i will suddenly lose the highs if i move my head or move around the chairs, or does it also have other effects like interaction with the room? or is it also something that have an effect even on axis?

[jmpsmash]

Hi Ian, lots of great info. Let me read through them in detail and understand/digest them one by one slowly and I will reply. Thanks!

[Ian Mackenzie]

There is no real time line with diy Audio.

What matters is you complete your project and achieve the desired outcomes.

I encourage you to be hands on and do your own investigations. That way you are more likely to achieve your desired outcomes and you will learn as you go.

[1audiohack]

Hi JMP;

I would skip the 2235's and buy 2216Nd's They are suprisingly inexpensive for a JBL Pro woofer that out performs the 2235 by a wide margin.

I love old JBL stuff and have 2234's and 2235's currently employed in systems I don't want to modify but when building a system like you are talking about from scratch the 2235's will leave a lot of performance on the table.

In truth a system like you are thinking, two 15" woofers per side will still be your SPL limiting factor as the 2123 and 2441's will run away from them easily.

Many people including me (if I have to) run the 2123's down to ≈250Hz but it doesn't play with real authority that low and 200Hz is as low as I ever use them. With really good 15's you can move that crossover up a bit and still keep the goodness the 2123 is known for.

The side by side wide baffle dual 15 JBL's like the 4350 and 4435 or DD66000/DD67000 have a good solid low frequency presentation that a single 15 just can't match.

I still think it takes front loaded bass horns to reproduce things like tympanies with true realism but you have to give up about 125 cubic feet of your space to do it right.

Just a note, my definition of "real" and "loud" are different than most people I know.

Barry.

[Ian Mackenzie]

One of your considerations will be how you stack the enclosures?

With 20 ft wall you will have either woofer enclosures upright or on the sides.
With that in mind you will want to space the horn centres between 8’ and 10’ feet) apart.

To start with your listen position (blast radius) will be 10 feet from each horn.

So the horns will be 5’ (feet) from each side wall.
The horns will want to be at what ever ear height you intend. Often 39 inches.

That will dictate the location of the drivers (ideally)

To avoid that wall of sound pa stack effect you will find clustering the horn / tweeter and woofer driver centres in a close vertical array important particularly if you enjoy classical music as opposed to over produced pop music.

Barry listens 50 yards from his system (apparently) because he still has his hearing.💭Lol

This is the problem with high powered systems consisting of multiple woofer enclosures and large horns. Alternatively you sit way back and hear your entire room along with the system. As l said investigate and work out with your enclosure modules what works for you.

[Dr.db]

What's the effect of having mismatched dispersion? is it for off axis, like i will suddenly lose the highs if i move my head or move around the chairs, or does it also have other effects like interaction with the room? or is it also something that have an effect even on axis?

What you actually hear sitting on your listening position is a mix of sound comming directly from your speakers and sound comming indirectly reflected from walls, ceiling and floor. That´s why a frequency response should be smooth on-axis AND off-axis. If your tweeter has a narrower dispersion than your midhorn you will have an uneven frequency response off-axis.
As explained above, you are listening to a sum of on-axis and off-axis sound in your room. In this case the highs would be overdamped.

Besides that, a sudden change in dispersion characteristics between individual drivers makes it easy to locate each sound source. The components dont match up well. What you hear are individual compontents, not speakersystem as one unit.

[Dr.db]

One of your considerations will be how you stack the enclosures?

The horns will want to be at what ever ear height you intend. Often 39 inches.

When using the 2397 horn and a supertweeter, how would you implement this idea?
The JBL 4343, 4344 or 4345 monitors mounted the 2405 tweeter right beside the midhorn. Both drivers were located on the same height / ear -height... But I´m afraid this can´t be done with a horn like the 2397!
This leaves you up to the decision if you put the 2397 or the tweeter on the ear-height, doesn´t it ?

[Ian Mackenzie]

I come back to my final points.

Don’t overthink the design.
The OP will need to do practical investigation.

A thread like this is only a discussion of suggestions and considerations. It is not a white paper about a loudspeaker design. All loudspeakers have compromises and it’s up to the OP to determine which compromises are acceptable to him.

Btw the DD67000 has the tweeter some distance above the horn. But what does the designer do with the knowledge of the particular characteristics of the horn?

You have to listen to the drivers to be prepared to do some meaningful measurements. None of this is beyond a diy audio enthusiast. With REW and a USB mic the tools are now accessible.

Another case in point is the approach Westlake took to their large format monitors.

The approach of a crossover point at 5000 hertz may actually be a good one depending on what the OP finds and prefers. I heard a horn system with a Fostex horn tweeter in the UK last year. It pays to have an open mind and do the work.

[jmpsmash]

I have received my Fostex tweeter and played around with it a little. The issue is that the 2397 is so deep and currently the Fostex and 2397 are driven by the same amp, just a simple cap on the Fostex to make the 5kHz cutoff. I need to get another DAC/amp to add the 4th channel.

First I put the Fostex at the front of the 2397. The 2397 being so thin there isn't much issue in terms of vertical height. That didn't sound too good image-wise with the time-misalignment.

Then I moved the Fostex back and the imaging was much more coherent and the sound stage moved back into the rear wall. The issue is that since it is sitting further back, there will be reflection bouncing off the top of the 2397. So maybe I need to raise it up a little. Will try to experiment with that later.

If I add a separate DAC/amp delay can be added and maybe it will work with the Fostex sitting at the front of the 2397. I will have to experiment with that too.

I need to do some measurement to make sure they are properly time aligned. Lots of work ahead.

Interesting that you mentioned DD67000. it has the tweeter horn also sitting back and there is a slope going down toward the lip of the mid-horn. Obviously they cannot add delay so maybe that's the best solution to get time-alignment.

[jmpsmash]

This is what I have cobbled together. More work to be done but I am already in love with the sound... well, anything 1kHz and above at least. :banghead:

85090

[jmpsmash]

Hi JMP;

I would skip the 2235's and buy 2216Nd's They are suprisingly inexpensive for a JBL Pro woofer that out performs the 2235 by a wide margin.

I love old JBL stuff and have 2234's and 2235's currently employed in systems I don't want to modify but when building a system like you are talking about from scratch the 2235's will leave a lot of performance on the table.

In truth a system like you are thinking, two 15" woofers per side will still be your SPL limiting factor as the 2123 and 2441's will run away from them easily.

Many people including me (if I have to) run the 2123's down to ≈250Hz but it doesn't play with real authority that low and 200Hz is as low as I ever use them. With really good 15's you can move that crossover up a bit and still keep the goodness the 2123 is known for.

The side by side wide baffle dual 15 JBL's like the 4350 and 4435 or DD66000/DD67000 have a good solid low frequency presentation that a single 15 just can't match.

I still think it takes front loaded bass horns to reproduce things like tympanies with true realism but you have to give up about 125 cubic feet of your space to do it right.

Just a note, my definition of "real" and "loud" are different than most people I know.

Barry.

Hi Barry,

you are really trying to "poison" me. Just 2 months ago I have visited a friend who has a pair of 15" and a pair of 18" on each side in some huge JBL pro cabinet. The amount of speed and ease coming from the low end was unbelievable and totally keeping up with the stack of horns sitting above. Unfortunately, I will have to get a divorce first before I can pull that one off. :D

This is the first time I have heard of the 2216ND-1. I did some research but so far they seem to be unobtainable anywhere. I can't find too much discussion on them, is it because they are less available? I would love to try them if I can find a pair. I always thought the 2235H is the best but wouldn't mind if there are better.

and let's not get into front loaded bass horns. where i live, the room itself will cost 100x the price of the horns... :banghead:

[jmpsmash]

Your ideas are good.

The gap is the acquisition of the 2235 woofers.

Talk to Edgewound at Upland Loudspeaker Repairs about getting old of some woofers or building up some woofers.

If you go with the dual 2235H woofers try them initially with the Smith Horn.

The Smith horns 2 inch throat limits diffraction up to 10,000 hertz.

Cut the Fostex horn over from 8000-14000 with a 3rd order filter. GT recommends a 3rd order high pass filter .

The Smith horn may not need a low pass filter.

Don’t over think the design. Start off with the basics and evolve following extensive listening.

Get the mini dsp USB mic and download REW for your initial measurements.

The 2397 should be smooth enough without any additional EQ.

Use a 12 dB LR filter on the woofer and horn @800 hertz (acoustic slopes) and either apply phase compensation by shifting the horn crossover point or adding delay to the woofers.

Don’t stress about the beaming of the woofers in the mid range too to start with.

Once you get this going you can then add the 2123H 10 inch mid cone.

Crossover at 250- 900 hertz with 12 dB LR (acoustic) slopes and apply phase compensation or delay.

As l mentioned you will need to measure the drivers individually and the crossover summation in order to get in the ball park. With some work you should be able to get a response +-3 dB from 35-20,000 hertz.

There are some good enclosure construction references in the Library and the technical references forum.

Use a suitable protection capacitor on the horn and the Fostex driver.

Tune the woofers enclosure to 30-34 hertz in 5 cu ft3

Enjoy your project.

Thanks. I plan to do some detail measurements of each individual drivers. I have obtained the Fostex, the 2241/2397 and the 2123H already. I still need to put together a cabinet for the 2123H. Right now all they set sitting on is just a front/open baffle and they make sound but not optimal.

I will read up on box design and building as well.

I am still looking for the 15". There are a couple pairs of 2235H on ebay. one pair has a mismatched looking cones and no surrounds, the other pair has really dodgy looking surround replacement job. both will end up around $800. Luckily I didn't decide on them as now Barry got me into looking for a pair of 2216ND-1

[1audiohack]

Speaker Exchange has them their site says. They are over $500 each which seems like a lot but when you total in core baskets and the shipping to get them to you, the cost of 2235 cone kits at about $200 each, install labor,, the shipping to and from a recone facility, you can easily approach $400 each by the time you get them ready to go. And they still are a dated design with OEM support weakening by the day.

Also any JBL PRO shop that can buy recone kits can buy the 2226Nd. The 2216Nd-1 seems a bit harder to get.

This is the woofer in the current top of the line JBL large format studio monitor, support there is as good as it gets.

Barry.

[Mr. Widget]

Speaker Exchange has them their site says. They are over $500 each which seems like a lot but when you total in core baskets and the shipping to get them to you, the cost of 2235 cone kits at about $200 each, install labor,, the shipping to and from a recone facility, you can easily approach $400 each by the time you get them ready to go. And they still are a dated design with OEM support weakening by the day.

Also any JBL PRO shop that can buy recone kits can buy the 2226Nd. The 2216Nd-1 seems a bit harder to get.

This is the woofer in the current top of the line JBL large format studio monitor, support there is as good as it gets.

Barry.You forgot to mention... they just plain sound good.
Great, is probably more like it! If I was going to do a DIY today, that is the woofer I'd go with.


Widget

[jmpsmash]

SpeakerExchange has both 2216ND and 2216ND-1 listed. the ND is in stock but ND-1 will take a couple of weeks. I just send them inquiry to see if they can actually get it and the price. Still googling all the reports. It seems that ND-1 is the better one.

[Dr.db]

I have never heard the 2216nd nor the 2216nd-1, but I believe they need additional eq in the lowbass? They have a much higher resonance-frequency compared to the 2235. It´s like 35hz vs. 20hz I think...
It seems they won´t deliever the rich lowbass as the classical 2235´s... Or will they perform similar with signal chaping?

[jmpsmash]

got reply from SpeakerExchange

2216ND-1 are $725 each
2216ND are $575

or I can get a used pair of 2235H for around $800-$900.

More research needed.

[jmpsmash]

plotted 2235H (red) vs. 2216ND-1 (green) on WinISD.

both are in 5 ft^3 cabs tuned to 28Hz.

In order to get the 2216ND-1 to achieve similar response, the cab will have to be 6.5+ ft^3

85094

[nedseg]

I paid around $800 (at speakerexchange) for my second ND-1, the first one (a year earlier) was priced just over $600.
First one took several weeks to arrive. Second one came faster, but arrived DOA, and was replaced within a week.
There are now quite a few comments about using these in upgrades or DIYs - posted here and on AudioKarma (and elsewhere).
Google is your Friend.
Alobar has posted his REW results on AK (Below), and I got essentially the same, or good down to 23hz or so, with only minimal EQ (I simply used just one of the lower DIY M2 EQ settings (POS, thank you) for the ND variant).
https://audiokarma.org/forums/index.php?threads/my-45-year-old-system-is-new-again.835545/
(This is in a 5 cu.ft. cabinet).
NDs or ND-1s, for me: Brand New, Under Warranty, deeper, cleaner, really deep bass, significantly improved lower midrange (~1khz) and imaging vs. used, older, 2235s of unknown history/condition (so needing service to be certain) for (very roughly) the same cost....the NDs were a no-brainer, and No Regrets!
That's my 2cents - but just enjoy the heck out of what ever route you take!
PS!
I also have a 4645C 18" (also with some EQ) - the 2216ND-1s will match it every step of the way (down), and more cleanly, but lose a tiny bit in sheer power (chest impact) delivered.
The 18" is driven by 1500w (Crown XLS), the ND1s by 150w (Bryston), too: so there's That, too.

https://audiokarma.org/forums/index.php?attachments/capture-jpg.1267567/

[jmpsmash]

I have never heard the 2216nd nor the 2216nd-1, but I believe they need additional eq in the lowbass? They have a much higher resonance-frequency compared to the 2235. It´s like 35hz vs. 20hz I think...
It seems they won´t deliever the rich lowbass as the classical 2235´s... Or will they perform similar with signal chaping?

that's the question. I am using DSP crossover which I can use to easily EQ the missing part. but then how much can one EQ? the missing dB compared to 2235H is very close to Fs already.

[jmpsmash]

I paid around $800 (at speakerexchange) for my second ND-1, the first one (a year earlier) was priced just over $600.
First one took several weeks to arrive. Second one came faster, but arrived DOA, and was replaced within a week.
There are now quite a few comments about using these in upgrades or DIYs - posted here and on AudioKarma (and elsewhere).
Google is your Friend.
Alobar has posted his REW results on AK (Below), and I got essentially the same, or good down to 23hz or so, with only minimal EQ (I simply used just one of the lower DIY M2 EQ settings (POS, thank you) for the ND variant).
https://audiokarma.org/forums/index.php?threads/my-45-year-old-system-is-new-again.835545/
(This is in a 5 cu.ft. cabinet).
NDs or ND-1s, for me: Brand New, Under Warranty, deeper, cleaner, really deep bass, significantly improved lower midrange (~1khz) and imaging vs. used, older, 2235s of unknown history/condition (so needing service to be certain) for (very roughly) the same cost....the NDs were a no-brainer, and No Regrets!
That's my 2cents - but just enjoy the heck out of what ever route you take!

https://audiokarma.org/forums/index.php?attachments/capture-jpg.1267567/

Very interesting! I agree witih you about risk of buying used. the price difference is not little but without have to do spend service cost.

Looks like I might have to give Speaker Exchange a call...

[Dr.db]

I wouldn`t pay 800$ for a used pair of 2235´s. Maybe for a freshly reconed pair...

To be honest I expected the difference in frequency-response to be bigger. The 2216 needs aprox. 2-3db eq at 30hz, that should work out. I guess JBL does this as well as the M2 is active....

Another contender could be the JBL 1500FE or 1501FE.... They could be a little cheaper and should be available at speakerexchange too. They have a nice 4" coil and very powerful motor. Maybe someone can compare these with the 2216´s??

[nedseg]

that's the question. I am using DSP crossover which I can use to easily EQ the missing part. but then how much can one EQ? the missing dB compared to 2235H is very close to Fs already.

From POS's DIY M2 EQ settings (see his sig), the biggey is ~ +5db @ 21.5hz.
I'm using a Venu 360 for xovers and room EQ, and really like it.
Look up the series of posts by NWCgrad which I found extremely helpful:
http://www.audioheritage.org/vbulletin/showthread.php?39133-Mostly-JBL-Pro-DIY-3-Way/page4

Happy readin' and spendin' !!

[Ian Mackenzie]

plotted 2235H (red) vs. 2216ND-1 (green) on WinISD.

both are in 5 ft^3 cabs tuned to 28Hz.

In order to get the 2216ND-1 to achieve similar response, the cab will have to be 6.5+ ft^3

85094

Stay in the 130-145 Lt and a tuning range of 27-34 hertz range and add an LF shelf boost to obtain the desired bass response.(2216nd). In a commercial product they can’t please everyone in every situation so they pick a compromise that
will cover most of the bases.

Again it comes down to doing some practical experimentation.

[Ian Mackenzie]

plotted 2235H (red) vs. 2216ND-1 (green) on WinISD.

both are in 5 ft^3 cabs tuned to 28Hz.

In order to get the 2216ND-1 to achieve similar response, the cab will have to be 6.5+ ft^3

85094

You will find building an enclosure and trialling it in your proposed listening room in different locations more meaningful.
One mans magazine curve can be a music lovers listening nightmare.

The 2235H is more extended on paper with that 28 hertz tuning. In a real listening space it will sound fatter in the 40-50 region as the enclosure is moved closer to a room boundary(s). Some people have become so acquainted with the thumping bass over four decades of listening they can’t live without it.

I know of other users who prefer the 2234H (2235H without the mass ring) and add modest bass boost as required.

There are a number of avenues to obtaining a favourable outcome. The key is to remain flexible in your thinking ��.
The final solution won’t be any one opinion or suggestion in this thread.

[jmpsmash]

Started doing all the basic work by measuring the 2397 and T925A.

The 2397 has a DSP 24dB highpass filter @ 950Hz, while the T925A has a 2uF cap in series, which should give it 6dB slope at 10kHz. They are connected to the same output, I just unplug one and plug the other one in. So the input supply is the same.

This was measured with the T925A sitting at the lip of the 2397. So the level don't really matter.

85104

For time alignment, I want to find out their impulse response. This is the part that bothers me. The impulse are 8 milliseconds apart. but physically, the 2 diagphrams are 20in (50cm). that makes no sense as 8ms means they should be 105 inches apart! Am I doing something wrong here?

85105

[jmpsmash]

Ok, I think I found out why. I can't compare the delay in separate sweeps. If I hook them up together, then I get this:

which give 1.4ms which end up 48cm which is close enough to what I measured physically.

So strictly, if I want to have them time aligned, the T925A need to sit on top of the 2441 body. That's way back.

85106

[jmpsmash]

And this is the combined response with the T925A sitting on top of the 2441 body. Looks not bad for a first pass!

JBL 2397 + Fostex T925A (w/ 2uF series). 24dB HP filter @ 950Hz

85107

[Ian Mackenzie]

On the basis your measurements are correct the issue is both drivers outputs should not be overlapping in the pass band.

The horn compression driver requires a low pass filter @5000 hertz. A 3rd or 4th order filter to attenuate the horn output above 5000 hertz. The Fostex should have an identical high pass filter @5000 hertz. You are best to do this actively.

This may require additional active crossovers.

Alternatively you could try a passive high pass filter on the Fostex at a higher frequency and use the horn drivers natural roll off as the low pass crossover. This will require a simple 3rd order high pass filter such as is commonly used with the Jbl 2405 slot radiator at about 9000 hertz. You may have already tried or considered this so l am suggesting this until you can do a 4 way fully active crossover. Typically this is what JBL does in this scenario with a horn system that requires additional HF extension.

You need to check the impedance of the Fostex but the parts values would be 1uF, 0.12mH, 1.5uF. I would place an 8 ohm Lpad on the Fostex and then connect the passive filter

[jmpsmash]

On the basis your measurements are correct the issue is both drivers outputs should not be overlapping in the pass band.

The horn compression driver requires a low pass filter @5000 hertz. A 3rd or 4th order filter to attenuate the horn output above 5000 hertz. The Fostex should have an identical high pass filter @5000 hertz. You are best to do this actively.

Yes. I was hoping I could avoid having to get another DAC/amp combo to drive the T925A, but from the results of this measurement, I do see that it is unavoidable.

1. I do see some interference between the two, esp if they are not aligned.
2. The placement of the T925A aligned with the 2441 is not ideal at all. I would like to put it further front which means adding delay to it via DSP.

[Ian Mackenzie]

In the short term you could improvise by moving the Fostex crossover point out to 9000 hertz and use a 3rd order high pass filter like Jbl used for the 2405 slot radiator.

The horn does not need a low pass filter as it is rolling off above 10000 hertz anyway.

I would use an Lpad just like in the schematic on the Fostex. This will provide a stable impedance for the passive crossover and allows you to control the level.

http://www.audioheritage.org/vbulletin/showthread.php?24032-New-4345-L-pad-wiring

This would be JBLs strategy in a situation where the horn system needed additional HF extension.

[Ian Mackenzie]

If you want to try this l can post the schematic.

To attempt a passive crossover at 5000 hertz for both the horn and the Fostex would be quite complicated in this diy situation and would cost you some $$$.

[jmpsmash]

just thinking aloud.

right now my plan is to go with the 2216ND-1, which is supposed to be very usable up to 1250Hz. I am planning to cross it at 1kHz with the 2397.

I currently have 3 DAC/amps driving the woofers, the 2123H and the 2397/Fostex.

Now if I can get away without the 2123H by using the 2216ND-1, then I free up one DAC/amp which I can use for the Fostex.

I read that it is better to cross the 2397 at 5-6kHz, which is my original plan, so I will have to use the DSP crossover anyway.

So in other words, going with the 2216ND-1 will save me having to pay for DAC/amp which means I save some money doing that. So I should stop delaying and go ahead and order them.

If that's the final solution, then I might as well save the money and not make any passive crossover. the 2uF I have lying around and it is good for some experiments.

That' the train of thoughts now at least. :)

[Ian Mackenzie]

If you are going the 2216ND-1 route that all makes sense.

Btw there is a lot of opinions across all this diy Audio space. One mans paradise is another mans nightmare!!
I used the 2397 for years with a 2405 passive crossover @8000 hertz and it worked great. But l had to customise the crossover to get a smooth transition.

You plan to use your existing Dac amp for the Fostex is good. You can try all these options and see what works best for you.

[Mr. Widget]

I read that it is better to cross the 2397 at 5-6kHz, which is my original plan, so I will have to use the DSP crossover anyway.You should verify this yourself. I think I’d go the 9KHz route myself.


Widget

[rusty jefferson]

FWIW, as I've never built a 4 way, but since you have them, I'd use the 2123 between the woofer and horn and live without the super tweeters until finances allow experimenting with the high frequency. You can get a lot of enjoyment out of the system running up to 10k. If you do cross at 5-6k those drivers will have to mesh perfectly. That's still a pretty critical range.

[dkalsi]

FWIW, as I've never built a 4 way, but since you have them, I'd use the 2123 between the woofer and horn and live without the super tweeters until finances allow experimenting with the high frequency. You can get a lot of enjoyment out of the system running up to 10k. If you do cross at 5-6k those drivers will have to mesh perfectly. That's still a pretty critical range.

100% agree with this!

[jmpsmash]

If you are going the 2216ND-1 route that all makes sense.

Btw there is a lot of opinions across all this diy Audio space. One mans paradise is another mans nightmare!!
I used the 2397 for years with a 2405 passive crossover @8000 hertz and it worked great. But l had to customise the crossover to get a smooth transition.

You plan to use your existing Dac amp for the Fostex is good. You can try all these options and see what works best for you.


You should verify this yourself. I think I’d go the 9KHz route myself.


Widget

Thanks guys. There will be a lot of experimentation for sure and I will post my findings here when I get to that point.

[jmpsmash]

FWIW, as I've never built a 4 way, but since you have them, I'd use the 2123 between the woofer and horn and live without the super tweeters until finances allow experimenting with the high frequency. You can get a lot of enjoyment out of the system running up to 10k. If you do cross at 5-6k those drivers will have to mesh perfectly. That's still a pretty critical range.


100% agree with this!

I have the 2397/2441 and the Fostex T925A already. Although not integrated well at the moment. Before I got the Fostex, I was listening to the 2397 on it's own. The sound is smooth, detailed, dynamic (esp micro dynamics). I really enjoyed the sound and didn't disappoint after the amount of work to acquire them. They sounded like what I remember them to be from 17 yrs ago.

They are excellent for vocal but they are missing the top end. At my age I am still able to hear up to 17kHz (thanks to the Fostex to verify that) so for something like a Jazz piece with the high hats and the upper end spectrum in classical pieces (flute, trumpet, oboes, triangles), they are missing something. Also the 2397 don't do placement very well. They don't paint a good image like cones and dome drivers. Adding the Fostex filled in that gap and give the whole presentation a lot more air and imaging.

I will have to play around and spend extra effort with driver integration.

[Mr. Widget]

Also the 2397 don't do placement very well. They don't paint a good image like cones and dome drivers. Adding the Fostex filled in that gap and give the whole presentation a lot more air and imaging.If you can control your first reflections, they can image pretty well. Unfortunately that 120° dispersion pattern is tough to deal with in most rooms.

I used 2397s and then the Westlake versions many years ago, and I believe I used a 3rd order passive network between the 2441s and the 2405s at around 9-10KHz. The tweeter was offset rearward some distance... unfortunately I don't really remember the details.


Widget

[Dr.db]

If you can control your first reflections, they can image pretty well. Unfortunately that 120° dispersion pattern is tough to deal with in most rooms.


I believe this is a major problem of this horn as just a few people have a controlled listening environment. Due to the size of these horns the speaker-enclosures get very wide and in average sized rooms they will be placed close to the sidewalls. This increases the problem with first-reflections further more...

I own the 2397 myself and love it´s sound a lot. But I believe in a small listening place a horn with 90° dispersion like the Yuichi A290 could work better...

[jmpsmash]

If you can control your first reflections, they can image pretty well. Unfortunately that 120° dispersion pattern is tough to deal with in most rooms.

I used 2397s and then the Westlake versions many years ago, and I believe I used a 3rd order passive network between the 2441s and the 2405s at around 9-10KHz. The tweeter was offset rearward some distance... unfortunately I don't really remember the details.


Widget


I believe this is a major problem of this horn as just a few people have a controlled listening environment. Due to the size of these horns the speaker-enclosures get very wide and in average sized rooms they will be placed close to the sidewalls. This increases the problem with first-reflections further more...

I own the 2397 myself and love it´s sound a lot. But I believe in a small listening place a horn with 90° dispersion like the Yuichi A290 could work better...

my room is 20ft wide. right now on the side walls there is a 2x4" diffusion panel on first reflection each side. it was there for my previous pair of regular dome speaker. let me make it 4x4 and see if that makes a difference.

[Mr. Widget]

my room is 20ft wide. right now on the side walls there is a 2x4" diffusion panel on first reflection each side. it was there for my previous pair of regular dome speaker. let me make it 4x4 and see if that makes a difference.My room was about 30’ by 25’... 10’ high heavy velvet curtains with a lot of “fullness” helped a lot.


Widget

[jmpsmash]

I was doing a bit of research on the 2216ND-1. the speaker that it was used in is the 4367. There is a nice a detailed review on it.

https://hometheaterhifi.com/reviews/speaker/floor-standing/jbl-4367-studio-monitor-loudspeaker-review/

the exterior speaker size is ~37x26x16.75. taking off 1" panel thickness and a bit more for the front baffle and that lip for the horn. I assume gross internal volume to be 5.5 cu ft or maybe slightly less. Not sure how much more to remove to account for the horn, woofer and bracing, probably another 1 cuft? so it will be down to 4.5 cu ft.

What I like about it is that it is wide and thin. The goal is to mount the 2216ND-1 as close as reasonable so if the cabinet is less than 17", the 2441 can dangle (supported) behind the back of the box.

There is also a nice internal diagram in the above article.

[jmpsmash]

alternative is to design it to be more acoustic ratio like, but the 2441 will either make the 2397 further away than the 2216ND, or I will have to put in a hole for the 2441 to sit in.

I plan to mount the 2216ND as high as possible within aesthetic constraints.

[jmpsmash]

I am trying to arrive at a alignment and thus box / port size for the 2216ND-1.

in a previous thread, @zeljkor measured a pair of 2216ND-1 to have:



T/S data measured from the first driver:
Re = 5.556Ohm
Fs = 28.26Hz
Qts = 0.376
Qes = 0.397
Qms = 7.12
Le = 0.461mH
Mms=134.1gr
Vas = 332.8liters

T/S data measured from the second driver:
Re = 5.52Ohm
Fs = 27.59Hz
Qts = 0.3572
Qes = 0.3776
Qms = 6.589
Le = 0.4703mH
Mms=123.8gr
Vas = 378.4liters


they differ a little as one is warmed.

I tried plugging into WinISD for a few different standard alignments.

SBB4 : 7.443 ft^3, tuned to 28.26Hz, F3=33.8, GD@40Hz=9.6ms, GD@30Hz=15.25ms
SC4: 8.869 ft^3, tuned to 29.48Hz, F3=30.5, GD@40Hz=10.6ms, GD@30Hz=18.5ms
QB3: 9.017 ft^3, tuned to 29.62Hz, F3=30.5, GD@40Hz=10.8ms, GD@30Hz=18.6ms

Those are pretty big boxes, but their FR is pretty flat with no hump. And supposedly they are the alignments with tighter bass.

QBB4 seems to be the most practical in terms of size.

What I still trying to understand is...

The 4367 cabinet is way smaller than the QBB4. I estimated it to be less than 5 ft^3. Is there something JBL knows? or is it just a marketing decision to go with a smaller box as a compromise?

SBB4 traded off 3Hz for a few seconds GD. I know 3Hz is probably audible, how about a few mseconds in GD?

There are also some stuffing in the 4367 cabinet. Is that something worth considering ? supposedly in closed box, it "adds" 10% effective volume.

What would be the sound quality difference if I go with a even smaller boxes? I know GD will go up which means slower and less tight bass?

How is everyone seem to be happy with just 5 ft^3?

I want to make sure the bass is tight good quality. If there is small sacrifice in F3, I can EQ it back. But how about a few ms in GD?

[grumpy]

I think the answers may lie between room gain and banana curve discussions.

[jmpsmash]

I think the answers may lie between room gain and banana curve discussions.

please explain...

[Ian Mackenzie]

In the diy space you have freedom to select your own enclosure alignment.

There are no cost, aesthetic or marketing constraints.

A properly tuned bass reflex enclose will only differ in bass extension and physically in enclosure size.

Generally the acceptance of a given box size is the controlling variable. You have to make that decision.

Once that is decided you can then manipulate the box tuning frequency to suit a preferred location in the room.

With the enclosure placed well away from a floor wall boundary intersection there is minimal boundary effect.

Moving the enclosure closer to the floor wall intersection will create a more pronounced bass response. Moving the enclosure closer to the corner of the room will further increase the bass output. The effect of the boundary placement needs to be assessed in a particular room for a particular loudspeaker.

The effect of a shift in the tuning frequency of 30 hertz over a small range (28-34 hertz) has a pronounced effect on the output of the system. This is because the woofer output tapers off significantly with frequency as the system approaches the tuning frequency. As this occurs the port output increases to a maximum at the box tuning frequency and then tapers off quickly below the tuning frequency.

As the tuning frequency is increased the bass output becomes more aggressive but the lower frequencies will diminish. A lower tuning frequency will produce a less aggressive bass output and will sound more damped.

With this in mind a skilled loudspeaker engineer can work with the box size and enclosure tuning so that as the enclosure is moved closer to the wall floor intersection the bass response output increases progressively at lower frequencies without becoming overly boomy in the mid to upper bass region. This is often an empirical process involving experimentation and listening tests.

Putting this another way a loudspeaker with a maximally flat enclosure alignment may look good on paper but may prove tricky when attempting to obtain a smooth extended bass response with the enclosure near a wall floor intersection.

This is because the mid bass frequencies will tend to dominate and appear louder than other bass frequencies due to a hump in the bass output with the enclosure placement near the wall. The bass quality can be sensitive to small changes in the distance of the enclosure from the rear wall. The user may then need to raise the enclosure off the floor to obtain the most satisfying bass response.

In comparison a sealed loudspeaker bass response will roll off at a higher frequency than a bass reflex loudspeaker but the rate or slope of the bass roll off is more gradual than a bass reflex design. Hence bass reflex designs have a reputation for being boom boxes in small rooms that don’t allow freedom of room placement.

Not every loudspeaker has a bass response that is going work in every room.

I therefore suggest you mock up an unfinished enclosure with a false back and try out some different enclosure volume and tuning options. Then experiment with room placement. You will more than likely prefer the bass of one enclosure and tuning over another in a particular room placement.

[Ian Mackenzie]

Another approach is to pick the minimum acceptable enclosure size for the driver and tune the system in accordance with a maximally flat alignment.

Then apply low frequency EQ using either a graphic equaliser or dsp equalisation if extension of the bass response is desired. This is a more straightforward approach than a customised tuning.

Generally a bass reflex enclosure has maximum power handling at the enclosure tuning frequency. Below that frequency the system is unloaded. You may therefore be able to modify the system low frequency response within reason. But be aware there are limits that need to be observed if overloading system the system is to be avoided. Moderation is important in how much EQ you apply as power dissipation and voice coil displacement must not be exceeded. A 3rd order high pass filter just below the tuning frequency is recommended if the system requires +6 dB of bass boost at the system tuning frequency.

If the bass response is dominant over other frequencies equalisation can be applied to reduce this effect.

Be aware that “cut” is preferred over boost in taming room related bass response problems. There are several reasons for this. Below 290 hertz room modes and other effects dominate the response in the form of peaks and dips. Attempting to boost a dip in the response may require significant equalisation and cause damage to the woofer. Dirac and other dsp based EQ is useful for cleaning up problems when flexibility in enclosure room placement does not exist. However not everyone likes the idea of converting an analogue source to dsp and then back to analogue. Analogue parametric equalisation is another option but can be expensive. The choice is yours.

Generally equalisation and adjustment to enclosure placement works most effectively when used together.

You can also consider room treatments. However such treatment is both expensive and bulky to be effective at bass frequencies. Also bad for WAF factor.

This is not an in depth discussion on the subject of loudspeaker and room equalisation but points to some of the key issues and choices available.

There is no benefit in aiming for perfection unless you are prepared to experiment with some ideas on a practical level before committing to a final design.

[jmpsmash]

Ian,

Thanks for the lengthy reply!

I like the idea of having a adjustable enclosure. I don't mind having a larger enclosure, I just want to have the best sounding one. So far I have only read about all these alignment and tunings on books and from forums and articles online. Even though I believe in theory, I also believe in hands on experimentation. Having such an enclosure will allow actual listening, experimentation and measurement.

My rough plan is to have a front baffle internal area of 18.5" wide x 30" tall. Adding 1 cu ft will be ~3 inches deep difference which is quite OK!

Wood will be cut this weekend.

[Ian Mackenzie]

Its entirely up to you what you end up doing.

However, with the 2397 as wide as it is (660mm or 26 inches wide ) I would shoot for an enclosure of that width.
http://www.jblpro.com/pub/obsolete/2397.pdf

That will give you more volume if you decide to go for a 7 cu ft enclosure.
Assuming your 30 inch (76cm) internal enclosure height then a 5 cuft3 enclosure will have internal depth of 11 inches and a 7 cuft3 will have an internal depth of 15 1/2 inches.

Remember a larger enclosure buys you a more extended bass response.

You can buy that bass extension in a smaller enclosure with these woofers by adding some EQ if required. Greg says you can hit the 2216nd with 8-10 db in a 25-27 hertz tuning in a 130 Lt enclosure.. This type of enclosure tuning delivers deep extended bass in a somewhat smaller enclosure. The trade off is amplifier power and lots off it. (reference to the JBL M2) As the voice coil heats up the DC resistance drops (due to the negative coefficient of the wire) and they punch hard. The 2216nd-1 cone has an acquaplas coating to lower the Fs a bit. Its otherwise a 2216nd driver.

As I said earlier when you EQ the system with your Mini dip you might find the EQ shaves back the bass response if you place the larger enclosures within 20 inches of a wall. Therefore the larger enclosure volume is wasted space. This is in part what JBL have opted for what appears a smaller enclosure than your diy box simulator suggests. The marketing department also knows sales will plummet with large enclosures (over 6 cuft3 net internal volume).

Once the room environment is introduced what you hear in the bass is the room influence not the best sounding enclosure tuning per say.
Your Minidsp EQ will clean up the bass.

I use a Nad C658 streaming preamp in a family room with a floor standing loudspeaker up against bookcases along the rear wall and the inbuilt Dirac EQ really improves the bass clarity.

By all means try a larger box of that is your desire but be careful what you wish for.

I hope this makes sense.

In the end you must decide what your going to do.

[Ian Mackenzie]

Its entirely up to you what you end up doing.

However, with the 2397 as wide as it is (660mm or 26 inches wide ) I would shoot for an external enclosure what wide.
http://www.jblpro.com/pub/obsolete/2397.pdf

That will give you more volume if you decide to go for a 7 cu ft enclosure.
Assuming your 30 inch (76cm) internal enclosure height then a 5 cuft3 enclosure will have internal depth of 11 inches and a 7 cuft3 will have an internal depth of 15 1/2 inches.

Remember a larger enclosure buys you a more extended bass response.

You can buy that bass extension in a smaller enclosure with these woofers by adding some EQ if required. Greg says you can hit the 2216nd with 8-10 db in a 25-27 hertz tuning in a 130 Lt enclosure.. This type of enclosure tuning delivers deep extended bass in a somewhat smaller enclosure. The trade off is amplifier power and lots off it. (reference to the JBL M2) As the voice coil heats up the DC resistance drops (due to the negative coefficient of the wire) and they punch hard. The 2216nd-1 has an acquaplas coating to lower the Fs a bit. Its otherwise a 2216nd driver.

As I said earlier when you EQ the system with your Minidsp you might find the EQ shaves back the bass response if you place the larger enclosures within 20 inches of a wall. Therefore the larger enclosure volume is wasted space. This is in part what JBL have opted for what appears a smaller enclosure than your diy box simulator suggests. The marketing department also knows sales will plummet with large enclosures (over 6 cuft3 net internal volume).

Once the room environment is introduced what you hear in the bass is the room influence not the best sounding enclosure tuning per say.
Your Minidsp EQ will clean up the bass.

I use a Nad C658 streaming preamp in a family room with a floor standing loudspeaker up against bookcases along the rear wall and the inbuilt Dirac EQ really improves the bass clarity.

By all means try a larger box of that is your desire but be careful what you wish for.

I hope this makes sense.

In the end you must decide what your going to do.

Edit. The Mms=134.1gr data you posted might not be accurate as the acquaplas coating adds mass to the cone (2216nd is 135gr).

[jmpsmash]

Again thanks for the very detailed explanation.

I plan to have the speakers around 5-6 ft off the front wall and 7-8 feet apart. That's where they (2397/2441 + T825A) are right now and they sound pretty good with a deep sound stage. I don't expect to move them that far into the front wall. I have a few things in mind:

having a 60cm baffle will push the baffle step cutoff down to 200Hz. I probably won't end up using the 2123H but would be nice for them to coincide with the 2123H low cutoff in case I do. I also not too sure about the look of a wider baffle honestly.

I have also in the back of my mind that, further in the future, probably some time next year, after I have figured out the crossover points, to build a passive crossover for the whole system. Why? Well to learn and gain some knowledge, and also to use a 300B amp that I have build but not utilized at the moment. I know it won't give the type of bass that my current amps will give, but something about claiming to have listened to something that I built by myself is very attractive. Also I am longing for the seductive sound of a class A SET plus horns.

[jmpsmash]

This is what I have so far. Still need to work on the bracing.

85158 85159

[Ian Mackenzie]

Your right it will image better out from the wall.

If you want to get fussy there are some interesting baffle diffraction modelling programs that predict the midrange response of the driver position on the baffle. Offsetting the driver a bit can smooth the response.

None of this matters if you use the Minidsp EQ but if you plan to use the SET solo it’s something you might consider.

If you do eventually try the SET amp my suggestion is Biamp with a high quality analogue active crossover so the SET is running the horn and the Fostex. You’ll get tight bass and that midrange you prefer. The mini dsp is a nice entry into active loudspeakers but not Hi end HiFi. Pm me when you want to explore analogue active crossovers.

My suggestion is to invest in the Dayton loudspeaker tester if you are serious about tuning your enclosures with actual TL data from your drivers.

https://www.parts-express.com/dayton-audio-dats-v3-computer-based-audio-component-test-system--390-807

Good luck with your project.😀

[jmpsmash]

A pair of this showed up in the mail today. :applaud:

I need that DATS but they don't sell the old one anymore and the new one isn't released yet!

85174

[jmpsmash]

Some progress over the weekend! Still need to finish the bracing and close up the back panel.

85199

[1audiohack]

Alright! Keep us up on this please.

Barry.

[jmpsmash]

Hopefully they will sing for the first time tonight!

I still have a few planned bracings to put on, however, I have made the back removable so I will play with measurement before and after those bracings.

I know. I am just impatient.

[jmpsmash]

Argh!!!

I have put the speakers together, but one of the 2216ND-1 is DOA. No sound and ohmmeter shows no conductivity. Quite a setback for a supposedly exciting evening. Now I have to go through the process of getting it replaced.

[Mr. Widget]

That is a bummer... but look at the bight side. If it was half of a pair of unobtainium vintage Western Electric woofers it would be worse.
Good luck getting it replaced with minimal drama.


Widget

[jmpsmash]

lol. it definitely sucks. now I can only listen to one side.

However, I do like what I am hearing. Very well defined bass. It goes down deep and I can not only hear but feel the bass.

I also start to realize I need to complete installing the bracing that I have cut, and potentially add more as I can feel and hear the box vibration.

[nedseg]

lol. it definitely sucks. now I can only listen to one side.

However, I do like what I am hearing. Very well defined bass. It goes down deep and I can not only hear but feel the bass.

I also start to realize I need to complete installing the bracing that I have cut, and potentially add more as I can feel and hear the box vibration.

I lived with only one ND-1 for a year before I could afford the second one (which also arrived DOA), but it gave me lots of time to compare (!) the vintage 2216 with the new tech version, and made me appreciate the 'totality' the second one provided THAT much more! :applaud:
Those things dig crazy deep, and with startling precision!

Best wishes on your RMA!

[jmpsmash]

I lived with only one ND-1 for a year before I could afford the second one (which also arrived DOA), but it gave me lots of time to compare (!) the vintage 2216 with the new tech version, and made me appreciate the 'totality' the second one provided THAT much more! :applaud:
Those things dig crazy deep, and with startling precision!

Best wishes on your RMA!

waiting a year will drive me mad!

right now I have routed both L/R channel to the one working speaker via the DSP.

keeping my fingers crossed the replacement will be smooth.

[SEAWOLF97]

Argh!!!

I have put the speakers together, but one of the 2216ND-1 is DOA. No sound and ohmmeter shows no conductivity. Quite a setback for a supposedly exciting evening. Now I have to go through the process of getting it replaced.

did you measure across the tinsel leads or just the post connections ?

[jmpsmash]

did you measure across the tinsel leads or just the post connections ?

there are no tinsel leads. the layout of the connection is a bit different than normal drivers. there is a connector plate at the back of the speakers and the wires go straight into the rear of the magnet structure, all sealed.

[jmpsmash]

lots of stuff todo.

- add more bracing
- mount the binding posts
- find a better solution to support the 2397/2441/T925A
- tuning
- decide on a finish. veneer or stain

85208

[macsic]

lots of stuff todo.

- find a better solution to support the 2397/2441/T925A


Hi,
It's this simple support that I made to support my 2397 and 2450. This year the 2450 have replaced my 2441 advantageously but it worked too.

https://nsa40.casimages.com/img/2019/10/24/191024031526504622.jpg (https://www.casimages.com/i/191024031526504622.jpg.html)

https://nsa40.casimages.com/img/2019/10/24/191024031526786074.jpg (https://www.casimages.com/i/191024031526786074.jpg.html)

[jmpsmash]

Hi,
It's this simple support that I made to support my 2397 and 2450. This year the 2450 have replaced my 2441 advantageously but it worked too.





macsic, very nice! I was planning to just 3D print something.

How did you find the 2450 vs 2441?

[jmpsmash]

update on the 2216ND-1. SpeakerExchange has contacted JBL. I am waiting for a shipping label to send JBL the DOA driver and they will send one to me directly. Keeping my fingers crossed.

[Dr.db]

Argh!!!

I have put the speakers together, but one of the 2216ND-1 is DOA. No sound and ohmmeter shows no conductivity. Quite a setback for a supposedly exciting evening. Now I have to go through the process of getting it replaced.


Unfortunately I have noticed quality issues with current JBL parts quit often in the near past. I don´t know if I would buy current JBL products anymore due to bad quality... Very sad to say, but JBL is still asking top dollars as allways but can´t remain it´s good quality from the past...

[jmpsmash]

Unfortunately I have noticed quality issues with current JBL parts quit often in the near past. I don´t know if I would buy current JBL products anymore due to bad quality... Very sad to say, but JBL is still asking top dollars as allways but can´t remain it´s good quality from the past...

One would think just checking voice coil resistance should be a simple and mandatory QC. Hopefully the one that works, will keep on working. There won't be so many of these 2216ND-1 around in 10-20 yrs time if they fail in the future.

[1audiohack]

Just a question, did you by chance lift the terminal plate off the back to see if the wires are connected?

It’s not like I have bought enough of these to statistically qualify anything but I am eight for eight on the 2216’s, all mine are good.

Barry.

[jmpsmash]

Just a question, did you by chance lift the terminal plate off the back to see if the wires are connected?

It’s not like I have bought enough of these to statistically qualify anything but I am eight for eight on the 2216’s, all mine are good.

Barry.

Yes. I did. I measured from the closet point on the wire that I can have access to, which is the crimp connection at the back of the terminal plate.

Unfortunately it just seems that I got unlucky.

[1audiohack]

Damn!

[jmpsmash]

update on the 2216ND-1. SpeakerExchange has contacted JBL. I am waiting for a shipping label to send JBL the DOA driver and they will send one to me directly. Keeping my fingers crossed.

Just shipped it out.

Somehow the warehouse for JBL is in Syracuse, NY? Since I am in CA, I was hoping it would be in Northridge. Now the turnaround will be 2 weeks.

[Mr. Widget]

Yeah, you’re a decade or two, too late. Of course back then your woofer would’ve weighed a lot more…, But it probably would’ve worked too. :banghead:


Widget

[jmpsmash]

Yeah, you’re a decade or two, too late. Of course back then your woofer would’ve weighed a lot more…, But it probably would’ve worked too. :banghead:


Widget

:banghead: indeed.

[macsic]

macsic, very nice! I was planning to just 3D print something.

How did you find the 2450 vs 2441?

I have always loved my 2441 until listening to the 2450 more "snappy" (how to say in english more alive) even better with Dia's Aquaplased than the 2441 (maybe too old). But you know it's my ears and those of my friends who arrive at this conclusion in my room with what I have.
The road you are pursuing is full of very pleasant discoveries and that is all the pleasure.
Good road to you :)

[jmpsmash]

making do with just one 2216ND-1 is hard! I did managed to do get quite a bit of bracing and then a little measurement done.

here is what it looks like right now. the back panel is attached with 20 screws.
85242

[jmpsmash]

Crossover points are 800Hz and 5kHz. 48db/octave LR

Frequency Response and Phase

85243

Distortion

85244

Group Delay

85245

[jmpsmash]

something is happening at around 200Hz. there is a null, a sudden phase crossing, and a increase in GD. that corresponds to the height of the cabinet...

I haven't added any damping or sound absorption in the box yet, just a braced bare wooden box.

[jmpsmash]

Also from the distortion, going from 2216ND-1 to 2397 at around 800Hz there is a sudden change/increase in distortion, same for going from 2397->Fostex at around 5kHz

In fact, one of the graph I made, shows that the 2216ND-1 can go up to 3kHz while keeping lower distortion than the horn!

there is also an increase in distortion going down from 50Hz. That might be the extra 5dB EQ I added there.

[jmpsmash]

Can't say too much about the sound at the moment. Since I am still waiting for the replacement 2216ND1. I only have bass on one side, and to compensate, I added a few dB to balance the sound level. Even then, it sounds unbalanced.

Before I added the bracing, I was able to hear some mid-frequency ringing. After spending a whole morning cutting and gluing a couple dozen braces, the box is much more quiet. Tapping sound dull instead of resonating.

I haven't done the damping yet. Decided to do some measurement first.

Any comments will be greatly appreciated!

[Mr. Widget]

something is happening at around 200Hz. there is a null, a sudden phase crossing, and a increase in GD. that corresponds to the height of the cabinet...

I haven't added any damping or sound absorption in the box yet, just a braced bare wooden box.Where is your measuring mic located? If it is not within a 1/4”-1/2” of the cone you will be getting some degree of room interaction. If you are taking a gated MLS measurement you need to raise the cabinet off the floor a meter or two in a very large room and take your measurement at about 2m.

If you don’t use close mic techniques or gated measurements you will definitely be getting misleading room influenced data.


Widget

[jmpsmash]

i did try gated but didn't make that much of a difference. However, I measured at 1m away. No way I am going to lift these up 1m, they weigh a ton! :banghead:

I will try close mic.

[Dr.db]

Also from the distortion, going from 2216ND-1 to 2397 at around 800Hz there is a sudden change/increase in distortion, same for going from 2397->Fostex at around 5kHz

In fact, one of the graph I made, shows that the 2216ND-1 can go up to 3kHz while keeping lower distortion than the horn!


How is that possible?
I would´ve thought that the horndriver should have way less distortion than the woofer.

[jmpsmash]

How is that possible?
I would´ve thought that the horndriver should have way less distortion than the woofer.

Yeah that's totally strange.

I can think of a few possibilities.

They are driven by different amps. I had first put the 2441/2397 on a pair of NuCore NC400. Mainly due to the low noise floor. the 2216ND-1 is driven by Nuprime ST-10.

They are also driven by different DACs. the 2441/2397 by a Hugo2, the 2216ND-1 by a Topping D70. I am pretty much scrambling for anything I have around. :)

I will try to arrange them so they use comparable electronics.

in either case, what bothers me still is that the FR for the 2441/2397 (800Hz to 5kHz) fluctuates much more than the 2216ND1 and Fostex. Here is the FR without smoothing:

85248

[jmpsmash]

this is also the 2441 that was giving me issue:

http://www.audioheritage.org/vbulletin/showthread.php?41885-2441-clipping-distorted-sound

I will try to swap in the other one and see. Maybe my alignment wasn't done properly.

[Ian Mackenzie]

Also from the distortion, going from 2216ND-1 to 2397 at around 800Hz there is a sudden change/increase in distortion, same for going from 2397->Fostex at around 5kHz

In fact, one of the graph I made, shows that the 2216ND-1 can go up to 3kHz while keeping lower distortion than the horn!

there is also an increase in distortion going down from 50Hz. That might be the extra 5dB EQ I added there.

Those results are consistent with the drivers of that era.

The 2216nd-1 is a very low distortion driver. If you do a ground plane measurement on your drive way it might be more revealing of the drivers actual response. Get some thin mdf sheets to form a smooth surface 2 metres square. Put the mic on the mdf and the 2397 at a distance of 2 metres. Do the same with the box but tilt slightly so the centre is aimed at the mic.

[1audiohack]

Well what’s one more opinion in the mix?

In my experience with measuring Smith horns the cells don’t play well with one another. If you run white noise through them and move side to side I can easily hear when I am in line with a vane or boat and they measure worse than they sound.

If you jam some reticulated foam in the cells adjacent to the center they measure much better (on axis with the open segment) via reduction of the cell to cell interference. That would eliminate much of your hash. If you put a mic just in it’s mouth on center you will see what I mean.

On the harmonic content, it seems to me that a driver cannot generate strong harmonics much beyond its passband unless it is grossly overdriven or it is damaged. The step up in harmonics at crossover don’t surprise me too much.

Fun project. :)
Barry.

[jmpsmash]

Well what’s one more opinion in the mix?

In my experience with measuring Smith horns the cells don’t play well with one another. If you run white noise through them and move side to side I can easily hear when I am in line with a vane or boat and they measure worse than they sound.

If you jam some reticulated foam in the cells adjacent to the center they measure much better (on axis with the open segment) via reduction of the cell to cell interference. That would eliminate much of your hash. If you put a mic just in it’s mouth on center you will see what I mean.

On the harmonic content, it seems to me that a driver cannot generate strong harmonics much beyond its passband unless it is grossly overdriven or it is damaged. The step up in harmonics at crossover don’t surprise me too much.


Is the imaging going to improve with the side cells blocked up? Something to test out.




Fun project. :)
Barry.

After going through the process of building, measuring and listening I am learning a lot!

[jmpsmash]

Those results are consistent with the drivers of that era.

The 2216nd-1 is a very low distortion driver. If you do a ground plane measurement on your drive way it might be more revealing of the drivers actual response. Get some thin mdf sheets to form a smooth surface 2 metres square. Put the mic on the mdf and the 2397 at a distance of 2 metres. Do the same with the box but tilt slightly so the centre is aimed at the mic.

Shame to say, I don't have a driveway or anywhere close to having open space in the near vicinity of my house that won't lead to me (and more importantly my newly built speaker) getting run over by a car... :banghead:

[pos]

Hello jmpsmash, fun project!
In your situation I think I would try something like this (https://reconingspeakers.com/product/jbl-eon-615-woofer-cover-plate-5039904/) to mitigate the differences in directivity between the 15" and the 2397, and potentially raise the crossover frequency a bit.

[1audiohack]

Did you say imaging and Smith horn in the same sentence? :)

Don’t go thinking I hate Smith horns, I don’t but imaging is not a character I attribute to them.

They are dispersive in the horizontal and diffractive in the vertical. Basically if not mounted in a large baffle they are sound everywhere and even in a baffle they blur the sound.

Some people like the diffuse sound everywhere and listening to “field” or large space recordings is where this type of reproduction kind of shines, in my opinion.

You can just cover the outer exits with a towel and they take on a much different character ie more direct but this just sends the energy back into the horn where reticulated foam or loose fiberglass will attenuate/absorb.

It might be worth playing with to see what you think.

Did I read you are crossing the Smith out about 5k?

I am not a Smith expert. I have a friend who went on a several year Smith tangent and I measured and listened to a bunch of variants of the design that he built with a wide array of drivers and throats. My opinions stem from this experience.

Barry

[Ian Mackenzie]

Indoor measurements for loudspeaker design are very limited. You can’t do meaningful power response measurements.

So don’t expect much.

If you have a car port try a ground plane measurement there.
You can’t do a near field measurements with that horn.

Do a gated mls measurement at 2 metres to get the 800 hertz crossover slopes as close as possible then reverse the phase to check the null. Then experiment with delay to adjust the vertical lobe till this have a uniform on axis response.
None of this is rocket science but the outcomes are only as good as your measurement data.

You might try 1 octave band pink noise either side of the crossover point to set the levels and test the null in reverse phase.

Your LF measurements can be near field.
Your distortion measurement if they are accurate don’t indicate a problem with the smith horn.
Use you mini dsp to knock out the room response bumps below 400 hertz.

Set up properly the 2397 has an open warm character but you will find large flat side wall surfaces a problem as far Imaging is concerned. The 90 x 40 Yuichi radial horns might be more successful.

[jmpsmash]

Hello jmpsmash, fun project!
In your situation I think I would try something like this (https://reconingspeakers.com/product/jbl-eon-615-woofer-cover-plate-5039904/) to mitigate the differences in directivity between the 15" and the 2397, and potentially raise the crossover frequency a bit.

with all due respects. I am not putting that thing any where near my speakers... :eek::eek::eek:

[jmpsmash]

Did you say imaging and Smith horn in the same sentence? :)

Don’t go thinking I hate Smith horns, I don’t but imaging is not a character I attribute to them.

They are dispersive in the horizontal and diffractive in the vertical. Basically if not mounted in a large baffle they are sound everywhere and even in a baffle they blur the sound.

Some people like the diffuse sound everywhere and listening to “field” or large space recordings is where this type of reproduction kind of shines, in my opinion.

You can just cover the outer exits with a towel and they take on a much different character ie more direct but this just sends the energy back into the horn where reticulated foam or loose fiberglass will attenuate/absorb.

It might be worth playing with to see what you think.

Did I read you are crossing the Smith out about 5k?

I am not a Smith expert. I have a friend who went on a several year Smith tangent and I measured and listened to a bunch of variants of the design that he built with a wide array of drivers and throats. My opinions stem from this experience.

Barry

I tried stuffing some acoustic eggcrate foam into the side cells and seems to tighten the image a little. Putting extra aborption panel on the side and front walls also help. I can get a better image but no where near what a small dome tweeter would do.

However, I also noticed the "sweet spot" is very narrow. moving my head just a couple of inches will spoil the image. I removed the Fostex from the chain and that's didn't change that effect. so that's purely from the 2397.

Surprising that adding the Fostex also don't change the imaging that much. I was trying to see if they are out of phase. They should be in phase according to the wiring but I wanted to hear from my ears. Unfortunately, the amount of information at 5kHz and beyond was very hard to discern. I was using one of those test CDs with in/out of phase voices. Can the phase information be deduce/measured from REW measurements?

[jmpsmash]

Indoor measurements for loudspeaker design are very limited. You can’t do meaningful power response measurements.

So don’t expect much.

If you have a car port try a ground plane measurement there.
You can’t do a near field measurements with that horn.

Do a gated mls measurement at 2 metres to get the 800 hertz crossover slopes as close as possible then reverse the phase to check the null. Then experiment with delay to adjust the vertical lobe till this have a uniform on axis response.
None of this is rocket science but the outcomes are only as good as your measurement data.

You might try 1 octave band pink noise either side of the crossover point to set the levels and test the null in reverse phase.

Your LF measurements can be near field.
Your distortion measurement if they are accurate don’t indicate a problem with the smith horn.
Use you mini dsp to knock out the room response bumps below 400 hertz.

Set up properly the 2397 has an open warm character but you will find large flat side wall surfaces a problem as far Imaging is concerned. The 90 x 40 Yuichi radial horns might be more successful.

Yuichi horns. every day I wake up and check this forum, I end up going more googling learning about new things. :D

Yuichi horns looks very interesting. also very involved construction. Maybe for a future project. I haven't been able to find any comparison with the 2397 though. What's the sound character? what I like about the smith horn is that it sound smoother than the regular horns.

A bit more update on the distortion. I opened up the 2441 again and went a bit further to align the diaphragm while playing piano and also sine wave to make sure there are no distortion at any frequency. I think I have it done properly this time. It ended up very late into the night so I ended up just sitting there listening to some music instead of doing measurements. Will do that tonight.

[1audiohack]

I also attest to the very good sound of the Yuichi A290. I have a pair of these that are currently on loan to my friend that has been on the Smith horn bent for years.

There is a guy in New Mexico that had some Yuichi kits he made for sale for pretty small money. Maybe worth a look.

In the comp driver front, some of those will give you a real run for your money. Some drivers and diaphragms just refuse to work together. Scatter those parts into different drivers and suddenly they're good do go. Stacked tolerances can be a nightmare.

On the foam front, if you can’t easily breath through it, you have the wrong foam.

Barry.

[pos]

with all due respects. I am not putting that thing any where near my speakers... :eek::eek::eek:
did not meant to offend you :p

[Dr.db]

I also attest to the very good sound of the Yuichi A290. I have a pair of these that are currently on loan to my friend that has been on the Smith horn bent for years.

Barry.

I would love to hear a comparison of both horns concerning the sound. :)
Of course the yuichi can be crossed over much lower than the 2397, but how do they compare above 1khz?

[jmpsmash]

I would love to hear a comparison of both horns concerning the sound. :)
Of course the yuichi can be crossed over much lower than the 2397, but how do they compare above 1khz?

I am ... *gulp*. very curious too. :D

[jmpsmash]

did some nearfield measurement of the 2216ND-1 and the port.

near field seems to be close to simulated.

the other 2 are the combined at 2 feet and 3 feet.

time to work on the placement and the room. :banghead:

85267

[Mr. Widget]

did some nearfield measurement of the 2216ND-1 and the port.

near field seems to be close to simulated.

the other 2 are the combined at 2 feet and 3 feet.

time to work on the placement and the room. :banghead:

Remember, listening trumps measuring. Trust your ears too.


Widget

[jmpsmash]

I spent some time playing around with the amps and looking at the distortion.

I ended up swapping out the NC400 and swapped in a single ended tube amp to drive the 2441 and another to separately drive the Fostex, and the distortion was more under control. make no sense as tubes are supposed to have high distortion (and there is still a 2nd harmonic dominant) and the NC400 is supposed to measure really well with very low distortion. Poor matching with horns for the NC400?

it also sounds way smoother with better holographic imaging. At that point, I was just enjoying the music.

I also played around with the crossover point. I tried 5kHz and 10kHz and found myself preferring 10kHz. Seems to paint a better coherent picture and sound more natural. Not sure if that's because the 5kHz wasn't tuned perfectly, or I just prefer the sound of the 2441/2397

[jmpsmash]

Remember, listening trumps measuring. Trust your ears too.


Widget

Thanks. that's great advise.

[jmpsmash]

some measurement and simulation definitely prove what Ian says. top is the measurement, and the bottom is the room simulation from REW.

having the actual simulation plus the near field measurement at least means the speakers themselves are behaving correctly. I just need to find a proper placement for them to sound their best.

85270

[jmpsmash]

And finally, the replacement 2216ND1 arrived and it works! A month after I ordered. Installed it and turning it up it is capable of some well defined thundering bass that shakes the room. Now I need to spend some time to figure out the right tuning for it.

[jmpsmash]

In the interim, I was trying out different configurations to match the 2441/2397 and the Fostex. trying out different amps, crossover points, active/passive, placements.

At the end I still like the sound of the Fostex sitting on top of the 2441. Physically time aligned. And then the fostex has a 2uF cap for a simple 1st order filter at around 10kHz. that's the original configuration. The sound is most coherent, the imaging is spot on, sound stage and wide and back into the wall.

maybe sometimes the most simple solution is the best?

[Mr. Widget]

maybe sometimes the most simple solution is the best?Almost always! :applaud:


Widget

[Ian Mackenzie]

In the interim, I was trying out different configurations to match the 2441/2397 and the Fostex. trying out different amps, crossover points, active/passive, placements.

At the end I still like the sound of the Fostex sitting on top of the 2441. Physically time aligned. And then the fostex has a 2uF cap for a simple 1st order filter at around 10kHz. that's the original configuration. The sound is most coherent, the imaging is spot on, sound stage and wide and back into the wall.

maybe sometimes the most simple solution is the best?

If you like it leave it as is.

[jmpsmash]

Almost always! :applaud:


Widget

:D

I think part of it is that they are now driven by the same amp and DAC. So there isn't a mismatch in sound between the drivers.

the upside is that one fewer DAC/Amp cost less money.

the downside is i prefer the Fostex to sit up front near the lip of the 2397. purely aesthetic, but hey... :bouncy:

[jmpsmash]

If you like it leave it as is.

I will do so and now that I have two woofers I will spend my time getting the 2397/2216ND-1 match better. I was having fun and really pumped up the bass and the 2216ND-1 handled it with ease.

[jmpsmash]

with a cabinet this large, is it realistic for it to be completely inert when the driver is pumped way high in volume? from touch alone, mine is not, there is some vibration, but at the same time, it is not resonating like crazy either. quite dull, but still there.

[jmpsmash]

Response at my listening position, with the speaker tuned to flat at the ear. There is a pretty big suck out below 50Hz and then recovers after 30Hz. I tried to simulate it in REW room sim and it confirms that suckout. I tried to do different things and none of which helps other than making all walls broadband perfect absorbers.

85300

Having mentioned, this flat response is rather... flat. After that, I tuned it to be closer to Harman curve which is way way more exciting to listen to. The amount of air it is capable of moving is pretty incredible. I also don't sense much in term of blending issue between the 2216ND-1 and 2441/2397. I ended the evening sitting in my room and enjoy some good music in speakers that I personally built. Bliss!

[Dr.db]

:D
the downside is i prefer the Fostex to sit up front near the lip of the 2397. purely aesthetic, but hey... :bouncy:

This is what I did with a 2397 and 2405 tweeter, maybe you like the idea.

85308

[Dr.db]

with a cabinet this large, is it realistic for it to be completely inert when the driver is pumped way high in volume? from touch alone, mine is not, there is some vibration, but at the same time, it is not resonating like crazy either. quite dull, but still there.

No, some kind of vibrations will always remain.
Of course you can reduce these vibrations by 90% or more, but not completly.
I had build concret enclosures with 1,5" panel thickness, even those vibrated a little.

A couple of years ago I did build these enclosures.
1,2" plywood with extensive bracing. These cabinets had 8 cubic feet internal volume.
Even with the bass playing at maximum the vibrations are barely noticeable! I wouldn´t put that much effort into enclosures again, but it really worked out great!

85309

85310

[Mr. Widget]

with a cabinet this large, is it realistic for it to be completely inert when the driver is pumped way high in volume? from touch alone, mine is not, there is some vibration, but at the same time, it is not resonating like crazy either. quite dull, but still there.Cabinet vibrations can be reduced significantly with rigid construction and added mass to absorb the excess energy. Westlake uses massive amounts of acoustic damping goop and their very large cabinets are very inert.

In my larger projects I have used a 3/4” MDF inner box or panel constrained by an outer layer of 3/4” birch plywood since it is much stiffer. I use massive internal bracing as well. I have not subjected this construction to any advanced analysis which would be beyond my rudimentary math skills, but subjectively it has been very successful.

Even at extreme SPLs the twin 2242 subs I built for Scott Fitlin for his disco system were non-resonant. My Project Widget speakers use this same construction except on the curved sides. There I have 4 layers of 3/4” MDF in the center of each panel. Running my hand on these panels at extreme SPLs, there is an extremely slight sense of vibration. The floor of the room and every other surface in the listening room is far more excited... is taking that on the next project? :D


Widget

[jmpsmash]

Dr.db & Mr.Widget,

those are some serious bracing you guys have done. And now that I have mine built, I understand why! In mine, I have mostly done bracing on each of the surface, sides and back have 4-5 2" brace, bottom and top have 3. front has corner and braces along the available spaces around the driver/port. I also have corner brace for the side/top/front. I don't have that many cross braces though.

anyhow, the box is pretty solid when knocked. But playing loud there are still vibration.

Next I need to figure out the padding inside. Right now I only have a cushion covering the back wall. I noticed the sound is more dull with it. Obviously not the correct implementation.

Previously I was also tweaking the 2397. I put some foam on the side channels. they do help the imaging, in fact, imaging is quite good. but it killed the broad sound stage and worse is that it killed the liveliness of the sound. So out the foam goes.

More work definitely need to be done there at the top end. If i take out the Fostex, it sounds even more natural. instrument timbre is more coherent, despite falling off above 10kHz. Adding the Fostex fills in the missing sound, but soften the texture. Maybe the Fostex is too smooth.

[Ian Mackenzie]

Standard practice is to start with 1 inch fibre glass insulation on all surfaces except the front baffle.
The fibreglass actually absorbs sound.

Do not use other materials such as carpet under felt.

Adding fibre glass will alter volume of air seen by the driver in the enclosure.
Adding 1 inch fibreglass adds approximately 10% volume to the enclosure. Too much fibreglass can create an over damped sound.

Attempting to make the enclosure completely inert acoustically is very difficult and may create more problems than it solves. For the purpose of identifying sound from the enclosure the way to think about it is to imagine your enclosure is mounted into a wall in a room. Your in the room behind the loudspeaker. How much sound can you hear coming from the enclosure? Assuming the wall is completely sound proof you will hear something. The sound from the box will ideally be Zero. Sounds are not always continuous tones. So you don’t want the box to resonate with a long decay time after exciting with enclosure with a short bursts of sound. Assuming the driver has an output of 100 dB you need to figure out an acceptable target for the enclosure? In the above scenario you could measure the difference. It might be 100 dB - 65 dB = 35 dB. It might be less or it might be more in total across all frequencies. You also have sound from inside the enclosure coming back out through the woofer cone. How do you stop that? You can’t. The fibreglass does however play a key role in absorbing sound at frequencies where this could become a problem. You can work towards a best case scenario by constructing the enclosure with golden ratio proportions, minimising panel resonances and other problems such as standing waves which focus at one particular frequency or set of frequencies. (See Golden Ratio). There is also sound other than bass frequencies from the port to consider.

My suggestion is don’t loose sleep over it.

My suggestion as discussed earlier is to create a 3rd order low pass filter on the 2397 and the Fostex at 8000-10000 hertz.

[jmpsmash]

do you just use the fiber glass insulation from the hardware store? is there a risk of the fiber getting airborne with all the vibration? The acoustic panel people are talking about how they are carcinogenic.

[jmpsmash]

Do not use other materials such as carpet under felt.


any reason why?

I have opened up a pair of former pretty top end speakers from a certain famous French speaker manufacturer and that's what they have inside.

[jmpsmash]

despite still some stuff to be done before I can call the speakers done (or will it ever be "done"? :) ) , the speakers are very very listenable and very enjoyable. great sound stage, great texture, incredible bass.

in fact, for the past few days, I have been spending a lot more time actually listening and enjoying music instead of busting my ass cutting wood, gluing, etc.

not losing track of what still need to be done. here is the list...

internal sound absorption
tune crossover points
finish - right now it is bare - it doesn't look bad at all but need to be more refined.

and something not directly the speakers, but to put in more acoustic treatment into the room.

[Ian Mackenzie]

any reason why?

I have opened up a pair of former pretty top end speakers from a certain famous French speaker manufacturer and that's what they have inside.

You can use other materials but they won’t have the same sound absorption coefficient as fibre glass. Fibreglass is a known quantity and it’s properties are predictable. Other materials may also reduce the volume of the internal enclosure. There are different sheet thicknesses and densities of fibre glass. Look at the Jbl enclosure guide in the Library to familiarise yourself with these recommendations. You can then decide on your options. Google is your friend.

[srm51555]

Awesome project. What was the final internal dimension? After reading your thread I finally got the ambition to try my 2397's with the 2216 and I will say they didn't disappoint, especially considering it was your standard Q&D horn on top of other cabinet. I was quite surprised on how the L300 charged coupled crossover (http://www.audioheritage.org/vbulletin/showthread.php?166-4333-S300-L300-equivalent-bandbass-circuit&p=1044&viewfull=1#post1044) on the 2440/2405's integrated with the standard M2 settings for the 2216. It wasn't absolutely perfect but pretty dang good. As Ian suggested run I'd the fostex in at 8-10K, no sense on waisting that sweet 4" JBL diaphragm's midrange. Keep up the good work, these will be keeper speakers.

Thanks,
Scott

[jmpsmash]

The internal dimension is 22.5x19x30.5. which bring it to 7.5 cu ft gross. net is approx 6.5 cu ft given all the bracing and the driver/port.

[jmpsmash]

I couldn't resist the temptation and got myself a pair of 2405H. Compared to the Fostex, it has more detail and more texture. However, the placement and sound stage is not as good. I have a harder time pin pointing instrument and voices, while the Fostex is much more precise. Tough choice between the two.

[Mr. Widget]

I couldn't resist the temptation and got myself a pair of 2405H. Compared to the Fostex, it has more detail and more texture. However, the placement and sound stage is not as good. I have a harder time pin pointing instrument and voices, while the Fostex is much more precise. Tough choice between the two.Can you try to absorb or defuse some of the early reflections? My guess is the beamier Fostex is imaging better for you because there is less reflected sound than you are getting with the much wider dispersion of the 2405.


Widget

[jmpsmash]

Can you try to absorb or defuse some of the early reflections? My guess is the beamier Fostex is imaging better for you because there is less reflected sound than you are getting with the much wider dispersion of the 2405.


Widget

I do have panels on the side walls on first reflection point. However, there is also quite a bit of clutter (who doesn't? :D ) around them. Perhaps with the wider dispersion I need to clean up the mess and make the two side walls more symmetrical. It will be a weekend project. :banghead:

[Dr.db]

I´m just curious if you still enjoy your system or if you have changed anything since your last post? :)

[jmpsmash]

I´m just curious if you still enjoy your system or if you have changed anything since your last post? :)

Well, long story. But audio is a journey, so long it is, bear with me. Since mid-November I got into an accident which resulted in a foot surgery to repair a broken bone. The accident was very much audio related but not for this pair of speakers... but that's another story :banghead:. Anyhow, I was relegated to crutches so while I am able to do small stuff it was work to move around, so not much work happened. later in December, vacationing mean I didn't get to do anything.

So when i was out vacationing, mom and I got really good front rows seat at Covent Garden and Teatro dell'Opera di Roma. I rarely sit in such prime location in concert hall with top acoustics and that was eye... i mean ear opening. The sound was large, massive, enveloping, quick, detailed, and mostly, not loud. Pizzicato from cello/double bass was quick, details, and while not loud, i can not only hear, but feel the sound. Natural, not piercing.

And most importantly, nothing like what I am hearing at home from the horns and the 15" woofer. At home sound was mostly too loud, when turned down to concert hall volume (I measured the dB at the hall, with my phone, I did... :D and matched it ), the sound was dull, compressed, no life. Sound stage was narrow, only between the speakers. Sound stage while still ok "deep", it wasn't massive. When loud, it was piercing. Details was ok. But overall, it sounded boxed in instead of this airy huge sound stage.

I couldn't get that sound off my head, and want to see what can be done. After I came back, I keep thinking and then remembered that last year I bought a pair of Heil AMT tweeters. So I start messing around with it. And it sounded... really good! The dipole tweeter engages the room, sound stage was bigger, more airy, more open. It is not as detailed as the horns but it was natural and not fatiguing.

I also posted a little in the other thread about it. But here is the story. I have been listening to and enjoying this new sound for the past week. It is not perfect, far from it, but at least from the sound stage point of view, I have not heard any non-dipole that sound like that.

I am still using the 2216ND-1 as the bass up to 1.2kHz, and it is really quite impressive as REW measures it matching the AMT in distortion at the crossover point.

Long way to go in this journey.

[rusty jefferson]

......The sound was large, massive, enveloping, quick, detailed, and mostly, not loud. Pizzicato from cello/double bass was quick, details, and while not loud, i can not only hear, but feel the sound. Natural, not piercing.

And most importantly, nothing like what I am hearing at home from the horns and the 15" woofer. At home sound was mostly too loud, when turned down to concert hall volume (I measured the dB at the hall, with my phone, I did... :D and matched it ), the sound was dull, compressed, no life. Sound stage was narrow, only between the speakers. Sound stage while still ok "deep", it wasn't massive. When loud, it was piercing. Details was ok. But overall, it sounded boxed in instead of this airy huge sound stage.

I couldn't get that sound off my head, and want to see what can be done......

Ah, the Audiophile's dilemma. The realistic dynamics of high efficiency horns (limited by their shortcomings), or the enveloping 3 dimensional recreation of a live performance from wide dispersion, time coherent loudspeakers (limited by their shortcomings). We should all have 2 listening rooms.:)

[jmpsmash]

Ah, the Audiophile's dilemma. The realistic dynamics of high efficiency horns (limited by their shortcomings), or the enveloping 3 dimensional recreation of a live performance from wide dispersion, time coherent loudspeakers (limited by their shortcomings). We should all have 2 listening rooms.:)

How about horns + full range driver/speaker pointing backwards?

[jmpsmash]

How about horns + full range driver/speaker pointing backwards?

Well thanks to MLK we have a 3-day long weekend and family is busy watching superbowl so I got the whole afternoon without interruption.

Took down the Heil and put back the 2397/Fostex. I use another channel of the DSP and wired up my old amp and speaker point it backwards towards the front corners.

I dug up some papers on concert hall acoustics and they referenced delay in the region of 100ms so I dial it to that. Initially I didn't quite like the sound, too loud and too bright, the highs makes it sound like everything are doubled. So I toned it down almost 10dB and tapered off the highs above 1kHz by another 3dB.

The effect is... fun. but can't say it is totally real. It does make one step towards that direction. The room is filled, there is direct sound stage coming from the main speakers, and then a field of sound enveloping the whole front of the room.

[1audiohack]

Just a crazy thought:

Point the horn forward and remove the cover from the HF driver and let that bounce around off the back wall. ?

Barry.

[Ian Mackenzie]

A lot of what you refer to in your comparison comes down to the recording and placement of your enclosures and the listening point also toe in.

It can take a while to get it right as it can be.

I start with an equilateral triangle and gruakually toe in the enclosures but not too much.

I tend to find getting the blend from the sides with the centre image balanced to your liking is what does it.

I also suggest you accurately measure and balance your L + R within 0.5 dB. You will find the sound scape snap right in once you do that.

The above comment are referenced to my 4345 in a room 4 x 7 metres fully furnished with heavy velvet curtains and 2.7 m ceiling. The enclosures are on the long wall.

As l said in is recording dependant . The Toy Yellow CD is an interesting recording but l find vinyl recording incl Miles Davis Kind is Blue and other classic vinyl albums superior in terms of that live / studio immersive listening experience.

I have done phase correction of the system and the differences were subtle in the vocals but the impact of AD was not satisfactory. I prefer tonal accuracy and purity and a front row presentation over attempting pin point soundstaging any day if the week. For example in a jazz bar in San Fran recently we had front row seats and it was incredibly emotionally to be close to the musicians. The intensity and tonal purity is what defines the experience for me. a really good system just brings you ever closer to that immersive experience.

I run a class A discrete analogue crossover specifically tuned to the voltage drive across the Vc of the drivers.

The result is a significant improvement over standard crossover setting and is superior to a digital crossover based on listening tests compared to the Mini DSP HD. For analogue addicts there are nothing better.

[Dr.db]

@jmpsmash: Thanks for those insights. :)




The Toy Yellow CD is an interesting recording
I wonder how you have found that CD :D

[jmpsmash]

Just a crazy thought:

Point the horn forward and remove the cover from the HF driver and let that bounce around off the back wall. ?

Barry.

That's a brilliant idea!

I will try it out and report back.

I did notice the horns are very piercing when compared to the Heil AMT. Mostly in voices which are 600-1000Hz. XO is 800Hz so that's right in the boundary. However, the harmonics are higher so perhaps the blending is not done very well, and human voice is harmonic rich.

I also tried pushing XO to 1200Hz and that's better.

Could be matching issues between the 2397 and 2216ND1.

[Robh3606]

That's a brilliant idea!

I will try it out and report back.

Hello jmpsmash


What kind of compensation do you have on the compression driver?? The "backwave" is going to be the voltage drive to the CD without the benefit of any horn loading so it may not sound right. If you are using a CD horn it may sound too bright because on the lower mid roll-off. Conversely if you are using and old style like an exponential, with no compensation, its going to roll off the highs at the mass break point and again without the benefit of the horn may sound dull up top kinda like a plain wave tube measurement.

You can't hurt anything by trying it out but just be mindful it may not match the horns output. And be really careful about any loose hardware you don't want a screw crashing into the diaphragm with the driver back off.

Curious to see how it works for you. I have toyed around with using an open compression driver as a driver but have never tried it.

Rob:)

[jmpsmash]

Hello jmpsmash


What kind of compensation do you have on the compression driver?? The "backwave" is going to be the voltage drive to the CD without the benefit of any horn loading so it may not sound right. If you are using a CD horn it may sound too bright because on the lower mid roll-off. Conversely if you are using and old style like an exponential, with no compensation, its going to roll off the highs at the mass break point and again without the benefit of the horn may sound dull up top kinda like a plain wave tube measurement.

You can't hurt anything by trying it out but just be mindful it may not match the horns output. And be really careful about any loose hardware you don't want a screw crashing into the diaphragm with the driver back off.

Curious to see how it works for you. I have toyed around with using an open compression driver as a driver but have never tried it.

Rob:)

That's a good point. I use DSP so I just tune it very slightly sloping down.

Yeah. I don't want to have the diaphragm exposed coz those things are unobtainable and cost an arm and a leg!

But i will try it anyway.

[Ian Mackenzie]

That's a brilliant idea!

I will try it out and report back.

I did notice the horns are very piercing when compared to the Heil AMT. Mostly in voices which are 600-1000Hz. XO is 800Hz so that's right in the boundary. However, the harmonics are higher so perhaps the blending is not done very well, and human voice is harmonic rich.

I also tried pushing XO to 1200Hz and that's better.

Could be matching issues between the 2397 and 2216ND1.

That is surprising for this Smith horn and 2440 compression driver.

I would re check your acoustic crossover points and slopes. Try 12 dB , 18 dB and 24 dB acoustic slope and up to 36 dB per octave back at 700- 800

Don’t be fooled into thinking that a 24 dB slope in the dsp is going to give you a 24 dB acoustic slope. You will need to assess with REW in each case manually. The horn and driver has its own native frequency response and phase response.
You don’t want to be adding more digital filters than you actually need. Less is alway more in dsp.

Then apply phase compensation on a needs basis and reverse the compression driver polarity to check you have a uniform null at 800 hertz. Measurements don’t always tell you what you are hearing.
You could be hearing the ragged edge of the 2216? Up at 1200 hertz and above.

On the imaging bear in mind it’s a pa horn designed to project into a large listening space. Your expectations might be beyond the component combinations you are using. Consider a 2311/2308 combo for better integration and a tighter dispersion aka the 4355. In comparison l personally preferred the 2307/2308 to the 2397. The 2311 should yield similar results when done properly.

Next l would run the setup using the drivers natural response without any dsp EQ response correction or room curve. Unless you have a very good understanding of the horn and room interaction leave it alone for now. The less processing the better. These components have been used for decades without any dsp diddle dee. In the pa world DSP rules with vast arrays and complex systems.

Your inside a relatively small space trying to make something better than listenable compared to a Walmart special. Otherwise what’s the point?
The sledgehammer approach is not appropriate when you have a lot of unknowns.

Dsp is not a workaround for problem solving in design of a loudspeaker system.


Greg Timber has recommended 18 dB slopes on the horn and the uhf device at 8000-10000 hertz generally.

I would stick with that recommendation.

Edit if you have a tilted response downward towards 20,000 this will highlight the mid range 600-1300 hertz. I would leave it natural and bring your uhf up in level to the point it adds crystal detail and air to complete the horns smooth midrange.

Always listen before asking the dsp to modify the responses.

[jmpsmash]

My best guess is the mismatch in dispersion between the 2397 and top end of the 2216ND1. I haven't done any off axis measurement for either. However, we know that 2397 go as wide as 160deg. I am betting that 2216ND1 narrow quite a bit at the high end. Since I only tuned them on axis, which may mean at crossover region, a lot more energy is projected into the room by the 2397 than the 2216ND1. My room has a decent amount of acoustic treatment with absorption at front and side wall. But still not gonna get rid of them.

Time to do some off axis measurements to confirm.

[pos]

I also tried pushing XO to 1200Hz and that's better.

Could be matching issues between the 2397 and 2216ND1.
My best guess is the mismatch in dispersion between the 2397 and top end of the 2216ND1.

http://www.audioheritage.org/vbulletin/showthread.php?41902-Building-my-dream-system-need-lots-of-advise!&p=425668&viewfull=1#post425668
;)

[Ian Mackenzie]

I agree with Barry’s comments on the 2397 application. You might try another horn as discussed much earlier in this thread. What l previously did was test and compare by ear 3 different horns.

I had previously used the 2397 with an Altec 515 @800 hertz. It worked okay.
I then added a 2405. Better.
Then l added an Audax hi sensitivity 8 inch mid driver from 250-1000 hertz. More improvements.
A pair of 2245’s replaced the Altec. Much better.
After l while l found the system struggling on classical music.
So l compared the 2397 to a 2344 and a 2307/23008. I preferred the 2307 with the 2405.
At that point a decided the clone a 4345. Far superior.

Building your own diy speaker is very challenging. Most likely you will fail or it might take years. I recommend you clone an existing Jbl system and spend more time listening to the music.

[Ian Mackenzie]

If you are proposing to continue with the Heil l suggest you email Nelson Pass at Passlabs on how best to set up the Heil. He previously worked in the R&D department at ESS when they introduced the Heil and build the crossover networks and has a wealth of expertise. He will respond.

[Robh3606]

If you are proposing to continue with the Heil l suggest you email Nelson Pass at Passlabs on how best to set up the Heil. He previously worked in the R&D department at ESS when they introduced the Heil and build the crossover networks and has a wealth of expertise. He will respond.

Cool didn't know that. They used 18db networks AMT-1 large Heil, The schematics are readilly available

http://esslabsusa.com/wp-content/uploads/2017/10/ESS_revisedfinal-2.pdf

Rob:)

[Ian Mackenzie]

They made the Heil drivers. Hint: There was a fair bit of QC on them to meet production.

[jmpsmash]

http://www.audioheritage.org/vbulletin/showthread.php?41902-Building-my-dream-system-need-lots-of-advise!&p=425668&viewfull=1#post425668
;)

:D Should've listened to expert's advises then? I knew much less then than now and it is great learning experience. It is one thing to learn from reading, another level to actually do it and "feel" it.



Building your own diy speaker is very challenging. Most likely you will fail or it might take years. I recommend you clone an existing Jbl system and spend more time listening to the music.


Cloning an existing one is different. It is like building from a kit. I don't run into issues and won't learn from all the design issues and decisions made.

As for listening to music, well, that happens at the end of the days when I am tired from measuring and tweaking. :bouncy:

[Ian Mackenzie]

Your taking the comments too literally.

You are progressing but don’t get bogged down on one approach.
Don’t try and pre conceive the design as perfect on paper. It won’t be.

You are right about the first bit.

Everyone wants to build their own loudspeaker.
But be careful what you wish for.

It takes quite a while to perfect a loudspeaker you can live with and enjoy.

If you are prepared to do that keep going and listen to the hints.
My example was four years. Buying expensive components is no guarantee of it sounding any good.

Building from a Jbl design.
It depends on your level of knowledge, skill, resources and the complexity of the design. Building an M2 clone is NOT a kit. There is a lot to learn from a Jbl design.

Building a known design gives certainty that you invested in something that really works (the outcome) and certainly you will finish. People miss sight of that and it all ends up being sold off on EBay if it doesn’t come together.

KISS Keep it simple if you are doing it on your own

Btw have you looked at the Linkwitz web site.
Those are excellent projects.

Or get hold of a pair of HPD 15 Tannoys and put them in Imperial horns. That’s used as a reference by one of the most successful American Hi end audio companies.

[Dr.db]

My best guess is the mismatch in dispersion between the 2397 and top end of the 2216ND1. I haven't done any off axis measurement for either. However, we know that 2397 go as wide as 160deg. I am betting that 2216ND1 narrow quite a bit at the high end. Since I only tuned them on axis, which may mean at crossover region, a lot more energy is projected into the room by the 2397 than the 2216ND1. My room has a decent amount of acoustic treatment with absorption at front and side wall. But still not gonna get rid of them.

Time to do some off axis measurements to confirm.

I believe this is a major factor.
If the off-axis dispersion drops significantly below the crossover, this could cause a thin sounding midrange. In the first instance we would blame the horn to sound harsh or thin, but it´s the midrange-woofer that doesn´t distribute enough midrange energy off-axis into the room.

Using a wide-dispersion midwoofer (8-10") below the 2397 horn should match the dispersion well.
When using a 15" woofer a horn with narrower dispersion than the 2397 horn should match better...

[jmpsmash]

Ok. I am convinced now that I should investigate further with the pair of 2123H that I have. Will do some dispersion measurement on those and compare with the 2216ND1 and see if that matches the 2397 better at 1kHz.

[Ian Mackenzie]

That combo works from own experience. It’s quite listenable. But l found it a double edge sword as the wider 2123 dispersion bounced off the 2397 lip causing a blurred sound (echo) on some program material. Any structure causes measured diffraction and reflection above 400 -500 hertz.

Nothing is perfect. It’s all about balancing the compromises.

This is a good thread on 2397.
http://www.audioheritage.org/vbulletin/showthread.php?3849-JBL-2397-Curves

Have fun.

[Dr.db]

@Ian:
I believe I have asked you before, but do you remember the distance between the top of the 2123 and the bottom of the 2397 horn back then?
I imagine it would need a fair bit of distance to keep the reflections low...

[Dr.db]

If you take a look at the westlake audio BBSM-15 it has the 2123 very close below the horn. Maybe 3-4"...
The BBSM-15 doesn´t use a 2397 horn, but the horn protrudes just as much.

On the other hand the westlake audio HR-1 has the 2123 mounted with much more distance to the horn. Maybe 8-10"....

[Ian Mackenzie]

It’s not a valid comparison.
The adjacent surface area of the Westlake horn in question is much smaller than the 2397.

[Ian Mackenzie]

Ok. I am convinced now that I should investigate further with the pair of 2123H that I have. Will do some dispersion measurement on those and compare with the 2216ND1 and see if that matches the 2397 better at 1kHz.


I noticed in one of your pics you were sitting the 2397 atop the enclosure.

This is common practice with audio amateurs.

Please be aware that Jbl provides baffle loading guidelines for the 2397.
My suggestion is to look up the 2397 data sheet to understand the guidelines.

Then cut a make shift loading baffle above below the 2397 and mount the baffles to the horn.
You might cut a slot in a sheet of 3/4 mdf to match the horn.

Then with the horn elevated at least 1.5 metres above any surface run an Frequency response and impedance curve and compare to the unloaded horn. Of course do a subjective test with an 800 hertz crossover point and 1200 hertz crossover point. You might find some of the issues you have had relate to the loading of the horn.

[jmpsmash]

I noticed in one of your pics you were sitting the 2397 atop the enclosure.

This is common practice with audio amateurs.

Please be aware that Jbl provides baffle loading guidelines for the 2397.
My suggestion is to look up the 2397 data sheet to understand the guidelines.

Then cut a make shift loading baffle above below the 2397 and mount the baffles to the horn.
You might cut a slot in a sheet of 3/4 mdf to match the horn.

Then with the horn elevated at least 1.5 metres above any surface run an Frequency response and impedance curve and compare to the unloaded horn. Of course do a subjective test with an 800 hertz crossover point and 1200 hertz crossover point. You might find some of the issues you have had relate to the loading of the horn.

The data sheet says:

"Because of the action of the power distributors, the 2397 is notably free of high frequency beaming, and maintains its effective 140deg horizontal coverage to the highest audible ranges, with only minor lobes appearing above 5kHz. Vertical dispersion is largely dependent on the baffling; with the recommended cylindrical baffle extending at least 3 inches above and below the mouth, a nominal 60deg vertical beam width is achieved. These dispersion characteristics make the 2397 especially useful for overhead mounting in rooms with low ceilings."

a cylindrical baffle will be quite a challenge!

[jmpsmash]

Ok. I am convinced now that I should investigate further with the pair of 2123H that I have. Will do some dispersion measurement on those and compare with the 2216ND1 and see if that matches the 2397 better at 1kHz.

Regarding that, a little bit of progress has been made.

I was doing some measurements and while setting up, happened to play some music with just the 2216ND1 on. And while measurements shows little in terms of distortion, the 2216ND1 played up to 1kHz area is ear piercing, not sure if that's due to the beaming, or distortion, or what, but totally killed my ears. No wonder the 4367 crossover is at 700Hz.

construction for the 2123H enclosure is done. ran out of time to fire it up though. will do that this week.

[Dr.db]

construction for the 2123H enclosure is done. ran out of time to fire it up though. will do that this week.

Looking forward to hear your impressions :)

[Ian Mackenzie]

I would put a 3 sided flat baffle above and below then try

[1audiohack]

I have heard very few systems that couldn’t be improved upon with the proper addition of a 2123. :p

Barry.

[Robh3606]

I have heard very few systems that couldn’t be improved upon with the proper addition of a 2123. :p

Yeah I really like those mega midrange drivers be it the 2108, 2122 or 2123. They can really add some snap in there!

Rob:)

[jmpsmash]

I have heard very few systems that couldn’t be improved upon with the proper addition of a 2123. :p

Barry.

I had to parse that a few times.... and still hope that's not what you meant. ;)

[jmpsmash]

I had to parse that a few times.... and still hope that's not what you meant. ;)

oh. pardon my poor comprehension of English. *NOW* I think (I hope) I understand what you meant.

Initially, I thought you meant:


I have heard very few systems that couldn’t be improved upon after the proper addition of a 2123

:o:

[1audiohack]

Sorry for being a goof.

I absolutely love the 2123. I am a dynamics nut and the clear reality the 2123 so easily delivers makes me want to use them in every system I build.

I have 10 more 2123H’s on the recone bench right now. :) It will be a sad day when these are no longer available.

Barry.

[pos]

Hey Barry, I'll take the opportunity to ask you something, as I trust your option/taste on those things.
I know you love both the 2020H and 2123H, but what about the 2012H then?

[jmpsmash]

Hi Barry, that's a great endorsement for the 2123. I got lucky to find a pair on local CL a couple of months ago. They were brand new sealed in box, the seller (a nice gentleman) won't cutoff the sealing tape before I committed to paying.

What in your experience would be considered a "proper addition" ? I plugged the TS param in and came up with a 0.5cu ft sealed box. I bumped it up to 0.67cu ft to account for the driver volume. I hope I am not that far off. It will be cut off around 250-300Hz at the bottom end. As for the top end, I will experiment with the 2397 before deciding, I have a whole lot more leeway there.

[Ian Mackenzie]

Yeah I really like those mega midrange drivers be it the 2108, 2122 or 2123. They can really add some snap in there!

Rob:)

They are all discontinued drivers now. The 2123 was used in numerous SR installations and they are now popping up readily on eBay.

I have heard them in rows of four in fixed installs with dual 2397’s and dual 15’s and it’s nice.

In the monitors you have to dial them each in with a good measurement set up and a simulator to make them work properly.

If you search the 2123 there are a number of threads on the correct sealed enclosures, Xmax limits and differences with the other 10’s. As to using them
Properly it requires the right tool and knowing how to use them. The limitation being most people use an RTA figure out what’s going on which does not have anything like the required resolution.

I have read posts claiming the 2216nd supersedes the application of there 2123. In a Jbl designed system l am sure they both have admirers. But often these claim comes from reading someone’s opinions without actually having heard either or ever having had any experience with either.

You really have to try these drivers yourself. Opinions are just opinions.

[1audiohack]

Hello Thomas!

You will laugh, I have never had any 2012’s. I look from time to time for them but haven’t bought any.

I recone just about everything I use and since JBL quality isn’t what it used to be in the voice coil department, I try to build at least six at a time and 10-12 if possible so I can match pairs.

I never see 2012’s as cores or in any quantity dang it.

Barry.

[1audiohack]

Hello jmp!

.5 to .65 cubic feet sealed filled with light loose fiberglass is the sweet spot for 2123’s when they are used like we are discussing as an upper mid bass / midrange.

While the 2123 will play lower than 250 Hz it doesn’t play with authority down there and I consider 250 the lower limit. With such a good woofer as the 2216 I would experiment with a little higher crossover to see what you like.

A friend and I have a project going together with 2216Nd’s, 2123H’s and 4338 horns with 2451SL’s and I am voicing a system with 2245’s, 2123’s and 4338 and 2451SL’s right now.

When we get to voicing the 2216/2123’s I will share what we end up with there. I expect these drivers to be excellent together.

Barry.

[jmpsmash]

Great. Looks like I am right on for the box alignment.

I think the 2216ND1 while can go up to 1kHz, it doesn't sound good up there. I look forward to hear what the 2123H can do.

As for the 2397 baffle. I have an idea how to do it. it will take a good weekend of work which means it will have to wait a little while.

I have to do more detail measurement. I meant to do the 2216ND1 dispersion measurement last weekend but was side tracked with other stuff.

I will test out the 2216ND1/2123H/2397/etc combo first, get a good sense of how they sound, and then start doing improvement to make it perfect.

This weekend's goal is to measure 2123H and measure 2216ND1. From that, I can figure out what crossover point will blend them well together, and then I can readjust with ears. Should be a fun weekend.

Barry, look forward to hearing your findings too.

[Dr.db]

When using the 2123 midbass, you propably end up with a little higher crossover on the 2397. I guess something between 1,0-1,2khz....
When using the 2216 woofer your crossover was around 0,8khz, right?

The additional baffle for the 2397 should be much more important with the lower crossover.
Maybe the higher crossover makes this additional baffle redundant...

[jmpsmash]

I think so. With the 2216ND1, the range is limited, I didn't like the sound of the 2216ND1 when pushed up to 1kHz, and 800Hz is stressing the 2397/2241 a little. With the more capable upper range of the 2123H, I have more option/headroom up in the 1kHz+ range.

Or to put it another way, more stuff to play with. :D

[Ian Mackenzie]

300-1200 hertz is ideal for the 2123H.

I would use the 2397 (vertical on axis) as the design point for the crossover and start with LR24 for the 2123H pass band.

You can apply delay on the 2123H to get the crossover point smooth on axis. But the off axis nulls are always going to be a compromise. I would measure the 2397 +-10 degrees off axis vertically to see what its doing at 2 metres. Same with the 2123H.

The 2123H relatively smooth but will require some shaping in the 800-1200 hertz region.

With the Fostex perched atop the 2397 your going to be sitting down to listen to this anyway. I would go with 7000-8000 hertz and 3rd order slopes for the 2397 and the Fostex. The 2441/2397 should be the dominant driver your listening to. The 2123H and the Fostex just clean up and top and bottom of the 2397.

If the 2397 still bothers you then you need to start looking at the Yuichi A290 or the FL version. The Tad TH4001 clones on Ebay are another option.

Edit
I suggest you keep this an all active affair and perhaps grab a minidsp HD 4 way module to get you started. If it all works out you can use the mini dsp settings for a better crossover later on.

[Dr.db]

How do you like those 2123´s ? :)

[jmpsmash]

so far so good. A bit more tamed than having the 2216ND1 do all the work. unfortunately life have been really busy and I haven't have much time to do listening and tuning. Will definitely put up more thoughts later.

[jmpsmash]

Finally have a little bit of time to work on the speakers. Here is the horizontal dispersion measurement for the 2123H.

85954

[jmpsmash]

and here is the one for the 2216ND1.
85955

[jmpsmash]

on the 2216ND1, how to decipher the narrowing dispersion and the dip between 150-300Hz? It comes back together and rapidly rising to 500-800Hz hump (which maybe what is bothering me), and then drops back down beyond 800Hz.

The 2123H is relatively better behaved, but looks best near the upper-200 and up to 1.2kHz.

[pos]

300Hz is a difficult frequency to deal with when measuring in a room, but with a combination of close mic measurement and spatial averaging around the listening position (eg MMM technique) it can be done.
I would advise you to use the same baffle width for the 15" and 10", to avoid baffle step differences and get a more (spacialy) consistent response so that you could ultimately rely almost only on close mic measurements.

[jmpsmash]

300Hz is a difficult frequency to deal with when measuring in a room, but with a combination of close mic measurement and spatial averaging around the listening position (eg MMM technique) it can be done.
I would advise you to use the same baffle width for the 15" and 10", to avoid baffle step differences and get a more (spacialy) consistent response so that you could ultimately rely almost only on close mic measurements.

good point about the baffle width/step. both the 2123H and 2216ND1 has the same baffle width, but that happens to fall around 250Hz. I am not sure if that's the issue though as the same drop don't happen in the 2123H graph. and then the 2216ND1 recovers ones it goes below 150Hz.

so maybe the baffle step being at around the same frequency is just a coincidence?

still don't understand what would cause it. could it be cabinet resonance of some sort?

[jmpsmash]

A bit of update on my project. After gotten distracted a little bit playing with my DAC. I came back to my speakers. A couple of things I did.

I flipped the bass cabinet sideways, this puts the 2397 closer to ear level, but don't look at nice as before and there is the mismatched baffle width. I also put back the 2405 instead of the Fostex.

Crossover is 250Hz, 1.2kHz, 10kHz

I have to say, now it sounds really good.

After swapping in the 2123H, I am no longer getting that beamy fatiguing sound from mid/high. I think that might be an issue with the 2216ND1 around 1kHz area. I cannot confirm though as I have not heard another implemention (ie. 4367). The 2123H does contribute to the whole stack in a positive way in that regards.

I didn't have a good solution to support the 2405 before so it was just resting on some wood pieces. I finally took the time to build a support for it.

I have to say that 2405 really shines here. Wider dispersion (the Fostex was very very narrow), but more important, more airy and a LOT of detail. I can hear more texture from almost everything. I did have to add a lot more room treatment to achieve focus, I think that helps with the super wide 2397 also.

I have been very much enjoying music since I arrived at this point. Now almost everything sounds good. It even revitalized those so-so orchestral recording.

Won't win any beauty contest, unfortunately. :)

https://photos.smugmug.com/photos/i-smCJ68N/0/8b296185/XL/i-smCJ68N-XL.jpg

[Robh3606]

Hello jmpsmash

Those are very nice! I really like the 2123. I use it in my HT system mains. I like what the JBL 10's bring to the mix.

I didn't know you had 2216nd-1's! What size box, tuned too??

How do you like them??

You have a 2440/41 on the smith horn??

I am really curious to see how the 2216's sound compared to my E-145's. I really like the E-145's, great snap and clarity.

If the 2216nd's are even better yahoooo!!!

Rob:)

[jmpsmash]

Hi Rob!

it's a (net) 6.5cuft net cabinet tuned to 30Hz. I like it. There is no replacement for displacement afterall. But I haven't compared it with other 15", at least not recently so I don't know for sure. It would be great if I can get my hands on a pair of 2235 and A/B compare! Without comparison, sometimes I wonder if I can get more definition out of it? or maybe the cabinet is too small/large? There is no official T/S param published for the 2216ND1. My room is also cursed when it comes to bass. it is a 20x20 ft room so I get a huge suckout at around 45-50Hz. Which literally sucks coz I am missing a lot. But that's how it is, I am lucky to have a room for music / hifi already. But in any case, right now they do sound very good! I have a few simple pieces where I have tried in various systems and config and the 2216ND1 produces the most defined bass on them.

yes, it is 2441 driving 2397 and the combination with the 2405 is really wonderful. Clarity, dynamics, presense, and with some work, I managed to get imaging out of them too (I have more than a dozen panels all over the room). Right now they are wired in parallel, using the natural drop off of the 2397 at 10kHz and I used a single cap on the 2405 to roll it off around that. For improvement I think I can do better with either another channel in my active crossover or build a proper passive.

[Ian Mackenzie]

That good real good.

Have a look at the 3107 schematic for the 2405.

I would follow the JBL 4367 on enclosure volume and tuning.

They added some acquaplas to the cone to lower the resonance but it didn't change it that much.

You could try measuring it with the DATS woofer tester of you want actual T/L parameters.

Aim for a 28-30 hertz tuning.

You might also try raising the woofer 4-6 inches off the floor.

[jmpsmash]

Ian, Thanks for the advise. I will look into the high pass. I have enough caps but will have to look for the air core inductor.

Giving the woofer a bit more listening. What I think is that it has enough volume and headroom. It can go very loud without any sign of strain at all. certainly enough to make me go deaf in my 20x20 ft room. However, it is def not the best bass I have heard. The best bass is not just loud and low distortion, but instead, fast, relaxed, and effortless. As if it is just strolling in the park.

I wonder if it is a issue with the enclosure? Or is it the amplification? I am using a Nuprime class D, 150wpc, 400 damping factor. I read a bit about Crown amps, do they have better control of the woofer? I see a Dci 2|300N for sale locally, 300wpc, >1000 damping factor. Will that be better?

The absolute best bass I have heard, from 2x15" + 2x18" per channel, totally fast, effortless, give me ear to ear grin, is driven by Crown amps. Was it the incredible amount of air moved, or the amps? Or... both?

[Ian Mackenzie]

Probably the woofer near the floor.

Try raising it off the floor for process of elimination.

You can generally tell with familiar program material if the tuning is out. It will either sound over damped or boom. If it’s sounds damped but blooms you need some PEQ below 150 hertz.

[Dr.db]

I guess the missing bass definition is caused by your square room.
A square room is horrible when it comes to room modes, as two modes are exactly at the same frequency and add up!
To be honest, it is very difficult/expensive to get it fixed with room acoustic treatments. If possible, move to another room.

[jmpsmash]

I guess the missing bass definition is caused by your square room.
A square room is horrible when it comes to room modes, as two modes are exactly at the same frequency and add up!
To be honest, it is very difficult/expensive to get it fixed with room acoustic treatments. If possible, move to another room.

I fully understand the physics and implications of the room dimensions. Unfortunately, with the real estate situation at my part of the world, I should be grateful to even have a room for hifi listening. I just have to deal with what I have.

Quite literally --> :banghead:

[Ian Mackenzie]

You might be able to moderate the problem with Minidsp EQ.

I had a similar problem in an apartment living room 20 years ago that was 4 m square and solid brick and concrete slab. The bass tended to hum. I found turning down the bass below 50 hertz on one channel helped.

[jmpsmash]

You might be able to moderate the problem with Minidsp EQ.

I had a similar problem in an apartment living room 20 years ago that was 4 m square and solid brick and concrete slab. The bass tended to hum. I found turning down the bass below 50 hertz on one channel helped.

I have a 10-15dB suckout. There is no way to fix the dip with EQ, it is a self cancellation. I lumped 9 inch 30"x48" think broad band panels in the corner and all they did was add 0.5-1dB. I would have to fill every wall corner of the whole room with that to get back that 10dB.

:banghead: :banghead: :banghead:

[Ian Mackenzie]

Wow,

That’s annoying?

Have you looked at it on REW?

Have you thought about renting a pad out in Palm Springs? Some of the modernist homes out there have substantial living spaces.

Honey we need change of pace.......
We loved it. The drive out of LA does suck though. The drive back to Santa Monica sucked more!

Send the wife out shopping with the other wives.

Then turn the room upside down.

Try offsetting the location of the enclosures so its asymmetrical.

Like moving them 30% into the room just to see what happens.

Stuff like that.

Also try moving the enclosures in or out from the wall and off the floor.

No room is every totally screwed if you can tolerate some changes.

I get some of the best sound when Madam President is out shopping ��!

Being her National Security Adviser l have a degree of influence like look at the Manley Stingray l bought you for your birthday darling...Lol��.

[jmpsmash]

lol. I am afraid my last name is not Zuckerberg nor Gates nor Bezos, otherwise I would have bought a prime piece of land and build my own custom room!

I do have my own room in my house that let me hide from the family when I need some me time to enjoy music, sometimes even late into the evening without bothering them. When my mom came to visit a couple of years ago (my mom is an audiophile and inspired my music passion and hifi hobby greatly), we sat and listened until 4am! I got lucky with having my own room, what I didn't get lucky is that the room is above the garage so it inherited a 2 car garage dimension which unfortunately falls exactly 20x20ft. :banghead: Win some, lose some.

Over the past year or so, I have accumulated more than a dozen panels including absorption and diffusion panels that I either purchased or built (I just made a set of 4 QRD13 panels last month!) and experimented a fair amount with it. I do make measurement with REW, esp for low frequencies. I don't do that much with high freq as most of my issues are imaging and echo/siblance and can't exactly measure that.

I can tame the high frequencies in my room pretty well. A few absorption panels and a few diffusion panels and it is really quite nice already. No glare, good imaging even with wide dispersion horns.

What I have learned is that to deal with bass, it takes a LOT more material. Small room can use a few bass traps, but for large room, I will need a huge amount of coverage. To achieve good coverage in a small room is ok. To do it in a 20x20 room, I will need meters and meters of very thick absorption. Even membrane based panels is the same. Paul of PSAudio tried to use Helmholtz. He built some into his room corners but he basically said it only gave him 1dB. I need 10db! I have seen youtube videos of professionally done installation when they line the whole side walls with Helmholtz. That's what it will take to tame the bass. Unfortunately I just cannot afford to do that. The sacrifice is too much.

I am sure there are solutions, I just need to do more research and consult more experience folks online to figure out how. But for now I have done enough with the room. I enjoy my time more working on the system first.

Having said that. While there is a hole in the response, I am enjoying the sound coming out of the system immensely. I believe while I cannot hear the sound, I can actually *feel* it.

[Dr.db]

It seems you have a lot of experience with acoustic panels allready.
You probably know this product allready, but if not have a look:
https://www.gikacoustics.com/product/gik-acoustics-scopus-tuned-bass-trap-t40/

I believe it would take at least 8-10 panels in your room.
But these should be very efficient, as they can be custom tuned to your specific frequency.

Another approach would be to ask a studio-designer to calculate a helmholtz-resonator for you. You can do the expensive part of building it yourself...
When finished, the studio-designer should visit for a last fine-tuning...

Both of these approaches will cost 2000-3000$, but if you can spend 1500$ on woofers this should be considered.

[jmpsmash]

It seems you have a lot of experience with acoustic panels allready.
You probably know this product allready, but if not have a look:
https://www.gikacoustics.com/product/gik-acoustics-scopus-tuned-bass-trap-t40/

I believe it would take at least 8-10 panels in your room.
But these should be very efficient, as they can be custom tuned to your specific frequency.

Another approach would be to ask a studio-designer to calculate a helmholtz-resonator for you. You can do the expensive part of building it yourself...
When finished, the studio-designer should visit for a last fine-tuning...

Both of these approaches will cost 2000-3000$, but if you can spend 1500$ on woofers this should be considered.

I did some research on both membrane / diaphragmatic absorber and Helmholtz. it seems that to go down to 50Hz they need to be in the region of more than 2m^2 each panel and 10" thick. There is a guy who build some room height ones with steel/rubber membrane but he never published the measurement. It look very involved. I don't know if I am prepared to have a few of those in my room at the moment. And my research kinda ended there as there are other things to play with...

As for GIK, they never published the data for their T40. The data presented have absorption peaked at 100Hz which indicates it is for their T100 panels.

If I do for it, the sacrifice will be quite a bit. Not only cost but space. It is a tough one.

[Ian Mackenzie]

What you could try is a local sub right near your listening position and inject the missing bass.

A local sub is referred to as global bass as l recall. It’s on top of you so to speak so there is little interference to be had.

I would simply place one of the enclosures right next to your seat as a trial with the other in the current position and see what happens. I imagine this would fill the hole you have.

If it works then it’s a much more viable thing to acquire a sub which is specific to that role.

Some subs like the SVS have very good Dsp that allows you to program what the sub does so integration should not be a barrier to making it work. Buy one with a 10 day return policy abs try it out.

I think you might like it.

[jmpsmash]

What you could try is a local sub right near your listening position and inject the missing bass.

A local sub is referred to as global bass as l recall. It’s on top of you so to speak so there is little interference to be had.

I would simply place one of the enclosures right next to your seat as a trial with the other in the current position and see what happens. I imagine this would fill the hole you have.

If it works then it’s a much more viable thing to acquire a sub which is specific to that role.

Some subs like the SVS have very good Dsp that allows you to program what the sub does so integration should not be a barrier to making it work. Buy one with a 10 day return policy abs try it out.

I think you might like it.

Yes. In fact, not right next to the chair, but 2 subs, one each to the left and right on the side wall will fill in all the holes. I actually modeled it in REW before. What remains to be found out is that, while that will fill in the FR, I wonder how that will sound? time to borrow 2 subs from somewhere. :)

[Ian Mackenzie]

I if you get a descent FR in the bass that is about all you need to know. Most subs have a phase adjustment

Beyond that do fall into the trap of trying to predict what might work or not work.

It comes down to some practical experimentation.

There are always going to be compromises that you have to accept.

You might actually prefer the sound with the subs as the 4367 is not known for its low bass extension.

[rusty jefferson]

Don't know what your situation is with your home but if it were possible to build a wall in your listening room to change the dimensions to approx 13'x20' if your ceiling is 8' high, or approx 14'x20' if 9' high, you'd be very close to the Golden rule for dimensions for optimal sound. Install a door in the wall and you'd still have access to the area for storage. Please disregard if your home is a rental. :)

[Mr. Widget]

Don't know what your situation is with your home but if it were possible to build a wall in your listening room to change the dimensions to approx 13'x20' if your ceiling is 8' high, or approx 14'x20' if 9' high, you'd be very close to the Golden rule for dimensions for optimal sound. Install a door in the wall and you'd still have access to the area for storage. Please disregard if your home is a rental. :)That is what I was thinking too. You can even incorporate soffits to hide the speakers.


Widget

[jmpsmash]

Don't know what your situation is with your home but if it were possible to build a wall in your listening room to change the dimensions to approx 13'x20' if your ceiling is 8' high, or approx 14'x20' if 9' high, you'd be very close to the Golden rule for dimensions for optimal sound. Install a door in the wall and you'd still have access to the area for storage. Please disregard if your home is a rental. :)

yes. unfortunately that's out of the question. :banghead:

[Dr.db]

I did some research on both membrane / diaphragmatic absorber and Helmholtz. it seems that to go down to 50Hz they need to be in the region of more than 2m^2 each panel and 10" thick.
The GIK Scopus is exactly 10" thick ;)
But they don`t have to be 2m² each panel, that might be true for membrane absorbers using steel plates. You will need several square meters of total absorption surface, that is correct!!




As for GIK, they never published the data for their T40. The data presented have absorption peaked at 100Hz which indicates it is for their T100 panels.


They have, here you go:

86320

86321

As you notice, in one picture the T40 peaks at 40hz, on the other picture at 50hz... I don´t know the reason, please contact GIK. But I have once phoned with them and you can have these specifically tuned! So measure with REW and let them build T40 panels with center-frequency of 38hz, 42hz, 47hz, etc.... Whatever your rooms asks for.

[jmpsmash]

Thankr Dr. db. The reason I say that is they have published official measurements by some lab for the T100. But such a measurement don't exist for the other 2 models. I do feel that at some point I need to give GIK a call... :banghead:

[jmpsmash]

It has been a while. Hope everyone are doing fine!

I have been playing around with my build for the past year.

I started off with active crossover. The first solution I went through, was a royal pain in the butt. It requires one DAC and amp per way and at one point I had 4 ways. I always think of it like it is on life support. So many things hooked up at the same time.

That led me into learning how to design passive crossover. I played around with XSim and tried out a few designs. Eventually I ended up with a design that sound not too bad. It is not where near perfect, but at least quite enjoyable and I don't need to hook up 10 pieces of equipment in order to listen to music.

I have learned a few things.

Firstly, I love the sound of the horns. I am not sure if it is specific to the 2441/2397/2405. But the details and microdynamic is quite addictive. Going back to dome tweeters and they just don't sound the same.

But then, with my current solution, 2441/2397/2405. Seems like a dead end. 2441 is so old, diaphragms are not available (I tried Radian, they don't sound the same, among other issues like I had problem aligning it while the original 2441 diaphragm I had no problems).

So I am thinking of going to a more modern route. I have been through many of the M2 thread, and the threads on newer JBL horn lens. The D2 driver seems interesting and they are still available brand new and cost not much more than a pair of nice dome tweeters. There are some WG available, the M2 horn lens, or STX825 horn lens. I have found members using it and also found some measurements.

What I noticed is that, despite having dual driver, the D2430K has huge drop in response that need to be EQ'ed. M2 has a active EQ, while 4367 passively compensates for it. Is that what's typical these days?

That's one part of my concern. I am ok to do that, since there will be a gap between the horn and 15" sensitivity so there is headroom for it. Or I will end up with a simpler 2 way active which is also OK.

My biggest concern is that, how will that sound compared to my current setup? Will I get the same type of sound? In the end I need to actually try it out in order to know, but if there is a consensus that it will be completely opposite, I might as well not try.

Thanks!

[Ian Mackenzie]

Have a talk to Rhob who recently did a passive monitor thread

I have not tried my M2 horns yet.

What you will find is that the passive compensation will require some tweaking to get it to your own liking. A flat response may not be ideal in your environment because these waveguides send the same output over a wide field into your room. So it might sound a bit too crisp or bright if you have it flat. So a mid tilt in the response above 1000 hertz maybe preferred.

The 2397 has wide coverage angle and a big open presentation but it falls off quickly above 10,000 hertz.

Using it in a 3 way with a 2405 you can manually modify the driver levels to adjust the tonal balance.

[jmpsmash]

Have a talk to Rhob who recently did a passive monitor thread

I have not tried my M2 horns yet.

What you will find is that the passive compensation will require some tweaking to get it to your own liking. A flat response may not be ideal in your environment because these waveguides send the same output over a wide field into your room. So it might sound a bit too crisp or bright if you have it flat. So a mid tilt in the response above 1000 hertz maybe preferred.

The 2397 has wide coverage angle and a big open presentation but it falls off quickly above 10,000 hertz.

Using it in a 3 way with a 2405 you can manually modify the driver levels to adjust the tonal balance.

Hi Ian,

The 4367 passive crossover was posted in one of the thread. I looked at it. One of the poles essentially started around 8kHz which compensates for the D2430K slope. Pretty wild but I will have to do something similar. I have been playing around with different orders for the passive of the 2405/2441/2216ND1 so hopefully it won't be hard to pull that off. I plan to play around with active crossover first and gradually migrate to passive.

As for the 2441/2397/2405. I really do like the sound. However, neither the 2441 nor the 2405 are perfectly matched. Not from measurement nor from listening.

My main concern is how the D2430K+STX825 will sound. I am hopeful that having a new pair they will be closer matched. But will they sound "better" or "worse"?

[Mr. Widget]

However, neither the 2441 nor the 2405 are perfectly matched. Not from measurement nor from listening.
Can you elaborate?

Widget

[jmpsmash]

Can you elaborate?

Widget

for both driver, both frequency response and impedance measurements, measured from REW and DATS respectively, don't match. Probably due to diaphragm getting old. I tried a pair of Radian 1246 and they measured fairly closely. Unfortunately the Radian has a huge dip at 1.5kHz and I had problem aligning the voice coil (didn't have problem with the 2441 original one).

[jmpsmash]

Have a talk to Rhob who recently did a passive monitor thread

I have not tried my M2 horns yet.


That's a great thread. Pretty close to what I am envisioning. I am gonna try the STX825 first, quite a bit more affordable to start.

[Ian Mackenzie]

Your just going to have to try it yourself

[jmpsmash]

Your just going to have to try it yourself

Yeah. I just placed order for D2430K + STX825. I am not the patience type. :applaud:

[Mr. Widget]

for both driver, both frequency response and impedance measurements, measured from REW and DATS respectively, don't match. Probably due to diaphragm getting old. I tried a pair of Radian 1246 and they measured fairly closely. Unfortunately the Radian has a huge dip at 1.5kHz and I had problem aligning the voice coil (didn't have problem with the 2441 original one).Got it. Yes, we often hear people's reviews of vintage speakers, drivers, and electronics... if you can't verify that they are up to spec, then you don't really know what you are listening to.

The effects of age on HF drivers in general can be tough and for these JBLs which often saw pro audio duty it is particularly difficult to find examples that are still up to spec, and finding a matched pair is even tougher.

In the early 2000s I was experimenting with a similar system to yours. I had factory rebuilt 2441s (yes, they used to do that :banghead:) and carefully rebuilt 2405s above 2202s and dual 2235Hs in a 4355 variant. They were impressive at high SPLs, but I never loved them at more reasonable levels and while I like the sound of the 2397, controlling that wide dispersion was problematic for me.

Going the M2 clone route is probably your best bet. The parts are all available new... not cheap, but based on my experience with JBL, if you don't buy them while they are available, you may be very disappointed in six months when they go on the NLA list.


Widget

[JeffW]

...and why anybody contemplating a build from new-ish JBL should invest in a backup set of parts. Last time I looked, even things like 2405 diaphragms were NLA, JBL has evidently quit supporting a lot of legacy equipment. If you build a system around any of the new-ish stuff that Harman didn't really want to sell in the first place, buy spares now.

[jmpsmash]

Got it. Yes, we often hear people's reviews of vintage speakers, drivers, and electronics... if you can't verify that they are up to spec, then you don't really know what you are listening to.


Agree. They sound very good, but probably a bit beyond their prime and not in their best performance. I do like the sound of the combo, so much detail, microdynamic and smooth at the same time. Many friends who came and listen to them have commented that they are ranked some of the top they've heard. But personally living with them for a long time all the imperfections sticks out more than the perfections. And of course, there is the urge to continue explore and learn about what else is out there. And esp something within the modern productions.




The effects of age on HF drivers in general can be tough and for these JBLs which often saw pro audio duty it is particularly difficult to find examples that are still up to spec, and finding a matched pair is even tougher.

In the early 2000s I was experimenting with a similar system to yours. I had factory rebuilt 2441s (yes, they used to do that :banghead:) and carefully rebuilt 2405s above 2202s and dual 2235Hs in a 4355 variant. They were impressive at high SPLs, but I never loved them at more reasonable levels and while I like the sound of the 2397, controlling that wide dispersion was problematic for me.


I think I understand what you are saying. I never was able to control it. And as a result there is certain disconnect and incoherence in the transition. That, and also the issue with a 2405 combined with 2441 transition. When they are far time aligned, there is no hope of phase alignment at crossover. And even when time aligned, it was still not smooth enough. I am expecting the D2430K solve that problem with dual diaphragm.



Going the M2 clone route is probably your best bet. The parts are all available new... not cheap, but based on my experience with JBL, if you don't buy them while they are available, you may be very disappointed in six months when they go on the NLA list.

Widget

great advise. I am still doing my research. plan is to try the STX825 first. see if I can work with the frequency variation. Then try the M2 one. Then I can decide which one to go with. I also learned a bit about the cabinet building and I think if I do decide to go with one of the two, I can build a new cabinet that will integrate the 2216ND1 and the horn into one neat package. Might end up with a 4367 clone, or M2 clone.

[Ian Mackenzie]

Greg Timber’s advice is to use a 3rd order electrical filter on the 2405 per 3144 schematic. I would look at the Jbl 3107 passive crossover for clues.

[Dr.db]

Maybe you can use some dusted 4" titanium - diaphragms in your horndrivers?

These should be available new for aprox 120$ and should be some kind similar to the 2441 type of sound.
That would remain the setup and crossover you have in general, but with fresh parts :)

[jmpsmash]

Greg Timber’s advice is to use a 3rd order electrical filter on the 2405 per 3144 schematic. I would look at the Jbl 3107 passive crossover for clues.

I modeled it in xsim and with some tweak they look pretty good. I will try to build it up to see how it sounds.

[jmpsmash]

Maybe you can use some dusted 4" titanium - diaphragms in your horndrivers?

These should be available new for aprox 120$ and should be some kind similar to the 2441 type of sound.
That would remain the setup and crossover you have in general, but with fresh parts :)

specifically which diaphragms are you referring to?

[JeffW]

specifically which diaphragms are you referring to?

D8R2450SL (https://reconingspeakers.com/product/jbl-d8r2450sl-diaphragm/) or D16R2451SL (https://reconingspeakers.com/product/jbl-d16r2451sl-diaphragm/)

They are calling the 16 ohm version discontinued, not sure if that's official or not.

[Kreativlos]

Honestly, if you're not satisfied with your current setup, it might be a design issue in general. Your System doesn't please you because the drivers aren't fine. They're quite quite good actually. So if you seek something else sound-wise, you may try a different Speaker Design with some of the key elements of your old design you liked. OR drastically rethink you design and the room, that surrounds it.
I would probably go for route 2, but I like the Design itself a lot. I'm biased tho, since I listen to a three way with Fostex components, but same idea. 12 inch bass ported, horizontally wide dispersion horn and so does the tweeter. I love it, and I'll probably never go another way.
I think you should try the route 1 and go for something like the M2, maybe with a super tweeter on top. It's the evolution of your old design and may be the right choice for you. You will still get plenty dynamic and detailed sound with smooth response and great micro dynamics. Thats just the Horn and big woofer style thing with quality drivers, the rest is about integration and how well the imaging is!

Cheers

[jmpsmash]

D8R2450SL (https://reconingspeakers.com/product/jbl-d8r2450sl-diaphragm/) or D16R2451SL (https://reconingspeakers.com/product/jbl-d16r2451sl-diaphragm/)

They are calling the 16 ohm version discontinued, not sure if that's official or not.

Interesting. The 2450SL is still available and not too expensive. Might try that.

[jmpsmash]

Honestly, if you're not satisfied with your current setup, it might be a design issue in general. Your System doesn't please you because the drivers aren't fine. They're quite quite good actually. So if you seek something else sound-wise, you may try a different Speaker Design with some of the key elements of your old design you liked. OR drastically rethink you design and the room, that surrounds it.
I would probably go for route 2, but I like the Design itself a lot. I'm biased tho, since I listen to a three way with Fostex components, but same idea. 12 inch bass ported, horizontally wide dispersion horn and so does the tweeter. I love it, and I'll probably never go another way.
I think you should try the route 1 and go for something like the M2, maybe with a super tweeter on top. It's the evolution of your old design and may be the right choice for you. You will still get plenty dynamic and detailed sound with smooth response and great micro dynamics. Thats just the Horn and big woofer style thing with quality drivers, the rest is about integration and how well the imaging is!

Cheers

As I mentioned in previous post. I am exploring STX825 horn with D2 (M2 driver). I am hoping it will give similar sound characters as the horns I have and have closer matching.

[jmpsmash]

Interesting. The 2450SL is still available and not too expensive. Might try that.

actually, take that back. same availability issue that I pointed out. shop in FL only have one 2450SL diaphragm left in stock and I haven't found any available elsewhere. And while it maybe easier to find (maybe?) given how many ppl have it, it might end up be used or from someone extra stash.

seems like approaching deadend if not already one.

[Ian Mackenzie]

Agree. They sound very good, but probably a bit beyond their prime and not in their best performance. I do like the sound of the combo, so much detail, microdynamic and smooth at the same time. Many friends who came and listen to them have commented that they are ranked some of the top they've heard. But personally living with them for a long time all the imperfections sticks out more than the perfections. And of course, there is the urge to continue explore and learn about what else is out there. And esp something within the modern productions.



I think I understand what you are saying. I never was able to control it. And as a result there is certain disconnect and incoherence in the transition. That, and also the issue with a 2405 combined with 2441 transition. When they are far time aligned, there is no hope of phase alignment at crossover. And even when time aligned, it was still not smooth enough. I am expecting the D2430K solve that problem with dual diaphragm.

great advise. I am still doing my research. plan is to try the STX825 first. see if I can work with the frequency variation. Then try the M2 one. Then I can decide which one to go with. I also learned a bit about the cabinet building and I think if I do decide to go with one of the two, I can build a new cabinet that will integrate the 2216ND1 and the horn into one neat package. Might end up with a 4367 clone, or M2 clone.

There is a significant time / phase discontinuity with the woofer and the 2397 compression driver abs the 2405. That can be a potential problem on certain types of program material.

I used the 2397 about 20 years ago and compared it to several other alternatives.

In the end the 2123 with a 2307/2308 and the 2405 had the least objectionable issues and l discovered Jbl made a similar design. It turned out very well and l still use it. Set up correctly and with well chosen amplifiers you will mer to spend over $10,000 on a current Jbl system to match it. There are some cool derivatives using large format drivers and some dusted diagrams that people speak highly off.

Sometimes in the search for the holy grail of diy loudspeaker you come across an engineered Jbl design that just works. Such a design is beyond the capabilities of diy audio amateurs because the original designer had access to professional resources and made a number of prototypes before settling on the final design.

Some of the more contemporary Jbl designs are a cool clone idea for the audio amateurs but these designs in some cars are voiced for the Japanese market where different tastes exist for local markets.

So there are no guarantees on what to expect with commercial any design or your own diy design.

There are people who have come and gone with the notion of buying the most expensive soa Jbl drivers only to find the drivers being sold off later because they either lost interest in getting it together or some other reason. Getting a loudspeaker together is not easy and not to be taken lightly.

[Kreativlos]

As I mentioned in previous post. I am exploring STX825 horn with D2 (M2 driver). I am hoping it will give similar sound characters as the horns I have and have closer matching.

Well, then good luck and please keep us up to date!:)

[jmpsmash]

I got a pair of D2430K and a STX825 horn. Being brand new, the matching of the 2 D2430K are spot on.

I built a little enclosure for the STX825 horn so that I can place it securely. The STX825 horn is smaller than I expected. They look out of place sitting on top of the 15" cabinet. Measurement of the combo shows nearly 15dB dip to 20kHz.

I started with playing around using passive crossover. It was tough work to compensate for the drop. I referenced the 4367 crossover topology and played with it. In the end, it was not possible to get rid of the 15dB dip and I still get 5dB drop after 10kHz. The sound was ok but lacking airy as a result. At one point, I got the 2405 on top of the D2430K and it filled in with a lot of airiness. the 2405 is very smooth indeed. However, the 2405 made the imaging worse.

After that, I went to active crossover. I have been playing with passive for the past year so had to dig out the minidsp box and do all the firmware update as there was a new version of Dirac which was much more robust and easier to use. The result is surprisingly good. The imaging is spot on, everything just falls into place with time domain correction. Instruments that are really difficult to focus like cello/bass/piano forms a image that I have not heard before. I was quite impressed. However, the combo doesn't have the smoothness of the 2405/2441/2397 combo. Sounds a bit more harsh.

Now I just got a pair of M2 horns. These things are so solid. I thought it will be just a piece of plastic like the STX825 but they seem to have filled it with resin. I originally thought it will ring if it was just plastic but this definitely won't.

I haven't mounted it yet so I will report back after that.

If it sounds right, I am tempted to build something similar to M2. So far I have built only speakers that look like DIY. I have learned a lot in the past 2 yrs not only in speaker design, but also in woodworking. In order to deal with the room mode in my room, I built 2 12" subs that are walnut veneered. I want to build a matching pair of speakers for the M2+2216ND1. The first design posted by Anti K in his KM2 thread has the K2 design which I think look really nice. I might do something in that direction, but not K2 half circle shape, it will be rounded walls but have to be deeper to get more volume. This will be a fun summer!

[jmpsmash]

I spent Saturday the past weekend listening to the STX825 horn. On Sunday I put together a temporary mounting for the M2 horn. Both using the same exact pair of D2430K. Unfortunately I only have one pair so I cannot A/B compare.

Using the D2430K alone with either STX825/M2 horn lens, and the same 2216ND1 below in a 2-way setup. There are a lot of similarities between the 2 but also a lot of differences.

The STX825 is more edgy, a bit sharper and harsher. However, it also means texture is a bit exaggerated. Normally with any other speakers, the violin section of the orchestra is just a glob of sound. The STX825 sounds like it is a group of instruments. Not to the point that one can point out each, it is not even possible in real life, but there is enough texture to not mush them up. The STX825 is really really sharp. Singers walking across stage can be very well pin point, and there are tons of space in between. However, that's at the expense of a bit too much harshness. At the end of the day, the ear can feel the fatigue.

The M2 is also sharp, but not as sharp. The overall presentation is a bit smoother, less edgy. Violin section is still detailed but not as much, singers are still separated but not pin point. However, I find the M2 horn to be easier on the ears and less fatigue.

That was the first impression.

[Anti K]

The first design posted by Anti K in his KM2 thread has the K2 design which I think look really nice. I might do something in that direction

👍😎