PDA

View Full Version : 2012H enclosure size?



Baron030
04-13-2005, 09:29 PM
Does anyone know the ideal enclosure volume and port dimensions for the JBL 2012H Midrange/Mid-Bass Transducer?

JBL's product sheet recommends 10-30 liters and an example of 8.5 liters tuned to 75 Hz.

What would the size of a 75 Hz port be?



I happen to be building a version of Daniel's Ancient Audiophile's quest for the Ultimate Home System.



So, any advice would be welcome.

Baron030
07-06-2005, 08:50 PM
Well, it has been several months now, and 60 people have chosen to view this thread.

But, unfortunately, no one has offered me any advice on the 2012 transducer. Since, I have about $500 tied up in a pair of these drivers. I have been forced to search else where for some sort of a workable solution. :(



Whenever, I have tried using an internet speaker enclosure calculator program, the resulting enclosure dimensions appeared to be way out of line with JBL’s recommendations. In both Bass reflex and Acoustic Suspension enclosure designs the internal volume appeared to be way too small. This has only added to my confusion about what to do with these drivers.:banghead:



Unfortunately, it would appear that this web site does not offer a lot advice on the 2012. But, I did come across a very helpful piece of advice when I widen my search to include “Dog Boxes”. Long time member “Giskard”, was helping someone with an enclosure question about the 2202H drive, when he stated the following:



Yes, but you might prefer the overall sound of ported instead of sealed. It's a preference and you should give both a listen to see which you prefer. The Qts of the 2202H is intentionally low so that any reasonable size enclosure (1.5 to 2.0 cubic feet) you put it in that is sealed will result in maximum transient response. That's what the JBL midrange transducers are designed for - maximum transient response in sealed enclosures. You might want to look on the Internet and read about Acoustic Suspension versus Infinite Baffle. The 2202H in a 1.5 to 2.0 cubic foot sealed volume behaves like an Infinite Baffle design whereas a transducer such as the LE10H in the same size volume behaves like an Acoustic Suspension design. Most JBL midrange transducers are intentionally designed to operate in sealed enclosures as Infinite Baffle systems to maximize transient response.



You can also play around with BB6P and see how different sized sealed volumes affect power handling, efficiency, response, group delay, etc. A Qtc of ~ 0.5 is critically damped, anything below that enters the realm of an Infinite Baffle design (the transducer begins to act as if it were in "free air" and the enclosure volume begins to have diminishing effect on the transducer). A Qtc of 0.577 is a Bessel response and a Qtc of 0.707 is a Butterworth response. It's probably best to stay away from anything higher than 0.707 in general.



Quote:

Originally posted by maxwedge

Is there a minimum depth that I need to be aware of that could cause problems?



Sure, your internal physical dimensions for any midrange enclosure should follow the "golden ratio" as midrange is the most critical and you want to avoid reinforcing internal standing waves like the black plague. There have been several discussions on this forum about the topic and numerous solutions worthy of investigation. Sonotubes, non-parallel sides, non-parallel baffle and back, etc.



I've really come to love those beasts.



Well, thanks to Giskard, I finally have an understanding of just how unusual the 2012 driver is. And that the ideal enclosure for this driver can be either an Infinite Baffle of 10 to 30 liters or a Bass Reflex enclosure of 8.5 liters tuned to 75 Hz. Since, mid-bass drivers are generally operated one to two octaves higher than their resonance, any differences in the frequency response between the enclosures shouldn’t be noticeable.



Currently, I have a pair of these drivers temporarily housed in Infinite Baffle enclosures with a net internal volume of 11 liters. But, I do plan on converting one of my test enclosures to a Bass Reflex design for a direct comparison.



Also, I may have found a simple solution for the 2012’s rising frequency response. If a 1.5mH coil is wired is series with the driver, the added reactance appears to flatten the response over the 400 to 1000 Hz range. Unfortunately, this simple solution will only work if you are using an active crossover and a separate amplifier for the 2012 drivers.

Hopefully, I will be testing 1.5mH coil equallizer and base reflex design soon.
And will let you know what the results are.

Baron 030

johnaec
07-06-2005, 09:04 PM
'Sorry I missed this. An 8.5 liter box tuned to 75 hz requires a 3" dia port 9.1" long. Since port velocity will probably not be a factor, you could also use a port of 5 sq.in. and 6" length for the same response.

John

stevem
07-06-2005, 09:40 PM
Baron, at what frequency are you trying to cross over? I also am trying to optimize these drivers, and in fact I'm in the process of building a pair of ported enclosures for them right now. I have found that in a sealed enclosure of about 30 liters, the response is down about 4 or 5 db at my crossover frequency of 300 Hz. Although I am also concerned about transient resonse, I'm hoping that reflex loading these drivers will flatten them at least at the crossover frequency. I have chosen to use a program suggested Vb of 8.8 liters tuned to 93 Hz. The higher tuning frequency should yield a flatter resopnse at 300 hz verses more extension at low frequencies. I should have them up and running by the end of this weekend. I'll let you know what I think then.

Baron030
07-06-2005, 10:06 PM
Hi John



Here is what I am going to be trying soon.

An enclosure with a net internal volume of 8.5 liters (0.3 cu’), works out to be about 9.5” x 12 7/8” x 6.25” is size. This is taking into account for the driver, 2 ports, and ¾” x ¾” cleats running along each inside edge. For the 2 ports, I am going to use 1 ½” Schedule 40 PCV pipe. The pipe has an inside diameter of 1.578125” and if my calculations are correct the correct length should be 4.9782” for 75 Hz. Oh, and as part of my calculation, I am assuming a 1 ¼” thick front panel. Hopefully, this should keep the port opening from coming too close to the rear panel. And believe me, It really is all a tight fit is such a small box.



I know this is going to sound strange. But the JBL Venue system uses a metal enclosure for this driver. It is to reduce power compression by dissipating heat out of the enclosure.

So, if I decide to go ahead with a bass reflex design, I am planning on using 2 ports.

One port will be mounted near the top and the other near the bottom of the enclosure. This will allow for a cooling convection current to flow through the enclosure.

And since, I will be using a Crown K1 to drive these 2012s with 350 watts per channel, heating may be come an issue.

GordonW
07-07-2005, 04:14 PM
Yeah, with the slightly rising response this driver exhibits (not to the degree of the 2123, but in the same realm), a higher-Q sealed enclosure (ie smaller) could definitely be used to "prop up" the bottom end response of the driver, to make it easier to crossover. It's quite easy to expect 300 Hz to be slightly raised, with an enclosure with a Q of 1 or so, compared to one with the traditional sealed enclosure Q=.707. It'll model showing a peak in the box-modelling programs, but IRL, once you add the effect of the rising natural driver response, it'll be much closer to flat...

As for vented- unless you're going to use it in such a way as to need serious bottom-end excursion out of the driver (ie, a 100 Hz crossover point, or something similarly low), I'd shy away from vented enclosures. It's always harder to passively cross-over a vented enclosure as a high-pass, than it is a sealed enclosure. That second impedence peak above box resonance is a PITA to deal with, using a passive highpass crossover in a ported midrange system...

Regards,
Gordon.

stevem
07-07-2005, 06:30 PM
As for vented- unless you're going to use it in such a way as to need serious bottom-end excursion out of the driver (ie, a 100 Hz crossover point, or something similarly low), I'd shy away from vented enclosures. It's always harder to passively cross-over a vented enclosure as a high-pass, than it is a sealed enclosure. That second impedence peak above box resonance is a PITA to deal with, using a passive highpass crossover in a ported midrange system...

Regards,
Gordon.

I'm using active crossovers. Do you think a vented enclosure would be a problem in that case? Maybe my sealed box is just too large (it's about 1 cubic foot).

GordonW
07-07-2005, 09:24 PM
I'm using active crossovers. Do you think a vented enclosure would be a problem in that case? Maybe my sealed box is just too large (it's about 1 cubic foot).

As long as the crossover point is AT LEAST 1.5 octaves AWAY from the port tuning frequency, it should work OK. Any closer, and the phase shift from the vented system rolloff, will combine drastically with the phase shift of the crossover. This will result in some rather wonky phase response... it'll make it harder to blend with the subwoofer/woofer driver. Sometimes, it's necessary to steepen (ie, use a higher order lowpass) the woofer crossover, to match the higher slope midbass acoustical total slope (electrical crossover plus box rolloff)... which can then straighten out the frequency response, but at the cost of group delay.

Point is, it's a matter of NOT wanting to stack up too many orders in the effective filters. Adding in that phase shift (and the associated group delay) isn't optimal in any case...

Regards,
Gordon.

GordonW
07-07-2005, 09:54 PM
Whoa.

Just ran the numbers on the 2012... I'd say, that box is too big!

One cubic foot gives a sealed box Qtc, of about .3. Now, under no conditions, is it beneficial to have a Qtc of less than .5... because not only do you lose bottom end extension, it also starts to ADD group delay in great proportions.

Truthfully, in a sealed enclosure, it's HARD to make the box TOO SMALL. If the dang driver will actually FIT in the enclosure, it's NOT too small. I mean, an air volume of .2 cubic feet (or about 5.6 liters), TOTAL, with box fill, would probably do very well... a Qtc of about .90. This will give a "plumped up" lower midrange/upper midbass which can counteract the rising response a bit. Box resonance would be around 270 Hz, which is where I'd put the crossover point. According to my calcs, the driver can handle full acoustic power at about there... a lower crossover in a sealed box, would be displacement-limited anyway.

If it were me, I'd at least try to make a test enclosure to try a very small volume for the 2012. For one example, a 10" sonotube round enclosure, about 5" deep (internal depth), would be a good one to try. Or, a 9" square (inside dimension) enclosure of about 4 3/4" internal depth. If you don't like it, then try some volumes somewhere in-between where you are now, and this suggested volume here. Somewhere in there, there should be a happy medium...

If you need the midbass to go lower than this, then I'd suggest a different midbass driver. The hassle of trying to make a ported enclosure mate up with a woofer, using an active crossover, is more than I'd personally want to live with. But, if you did want to try vented, the proposed 8.8 liter enclosure tuned to 93 Hz, is a decent option... it's probably about the most sensible compromise between low port frequency and "propping up" the midbass.

Regards,
Gordon.

stevem
07-08-2005, 09:01 AM
Thanks Gordon, I appreciate your comments. I'm actually trying to increase the output of the driver at the crossover frequency, that's why I want to give the ported box a go. I started with a .5 cu foot sealed box, and enlarged it to 1 cu foot to see if I could get a flatter low end response, but it really didn't do anything. When I run my box program with smaller enclosure sizes, I don't see an increase in the midbass response that you mentioned. Am I doing something wrong? I'm actively crossing over and EQing the HF rise out of the driver as it is. I'm really just trying to avoid boosting the response to flatten the falling resonse at the low end.

Baron, I feel like I hijacked your thread! Sorry! I hope you're getting something out of this too!

Miss Orchid
07-09-2005, 08:02 AM
Now, under no conditions, is it beneficial to have a Qtc of less than .5... because not only do you lose bottom end extension, it also starts to ADD group delay in great proportions.

Truthfully, in a sealed enclosure, it's HARD to make the box TOO SMALL.

There are JBL systems with Qtc numbers much lower than 0.5 and JBL thought that was beneficial. The readers of this forum are urged to examine group delay figures for sealed systems and vented systems in a variety of box sizes, and in the case of vented systems, a variety of tunings. Sealed boxes can be too small and can decrease power handling capacity due to thermal issues.

Also of interest might be the use of sealed boxes as two poles of a high pass filter.

I'm actively crossing over and EQing the HF rise out of the driver as it is. I'm really just trying to avoid boosting the response to flatten the falling resonse at the low end.

EQ the low end. Possibly a better solution than venting the box depending on the Quality of the EQ.

The sealed box is providing you with a 12 dB per octave roll off. How many poles in your active high pass filter for the 2012H. What happens when you lower the active portion of the total acoustic roll off.

When I run my box program with smaller enclosure sizes, I don't see an increase in the midbass response that you mentioned. Am I doing something wrong?

It's subtle. Read Baron030's second post above. What is the Qts of the 2012H. What is the Vas. Here is the Product sheet for reference - http://www.jblpro.com/pages/pub/components/2012h.pdf

.

stevem
07-09-2005, 09:52 AM
Thank you for your reply, Miss Orchid. It appears that I might be going in the wrong direction with ported midrange enclosures, but since the boxes are almost completed, I might as well finish them and see how they sound.



In answer to your question, I am using 48 db/octave Linkwitz-Riley crossover slopes, which I think are 8 pole? If I understand you, you are suggesting using the 12 db/octave natural roll-off of the sealed enclosure in conjunction with a 36 db/octave (6 pole?) slope for a 48 db/octave acoustical slope. Is this correct?



I don’t want to use a lot of EQ boost, because the digital crossover/EQ I’m using is running at full scale, and too much boost will make it clip. Global attenuation (in the digital domain) solves this problem, but it doesn’t seem to sound as good.



By the way, I’m very impressed with your knowledge of things JBL, as well as your command of the English language. You write like someone from the USA, like maybe Ohio?;) Anyway, thanks again!

Miss Orchid
07-09-2005, 10:30 AM
It appears that I might be going in the wrong direction with ported midrange enclosures, but since the boxes are almost completed, I might as well finish them and see how they sound.

Ok. It is good to try things.

In answer to your question, I am using 48 db/octave Linkwitz-Riley crossover slopes, which I think are 8 pole? If I understand you, you are suggesting using the 12 db/octave natural roll-off of the sealed enclosure in conjunction with a 36 db/octave (6 pole?) slope for a 48 db/octave acoustical slope. Is this correct?

Perhaps so. Perhaps lowering the active portion of the high pass to the 2012H will help.

By the way, I’m very impressed with your knowledge of things JBL, as well as your command of the English language. You write like someone from the USA, like maybe Ohio?

I command many things, how do you say, "grasshopper", including the English language. Know your enemy... And know this - China will crush your capitalist pig country. Oink! Oink! You have become an evil empire and must be stopped. We will buy all your corporations, snatch away all your jobs and smash your economy while you are busy spreading your McDonald's franchises all over the Middle East. Even now we are designing and building your precious JBL speakers. :rotfl:

Ohio? Ah yes! The place they shoot students who protest. We run over ours with tanks. Isn't that silly girl Katie Holmes from Ohio? Doctorate in Physics and Masters in Mathematics right? Oh wait, my notes are all over the place... she's the scientology major isn't she... yes, we know much about your Ohio...

Anyway, thanks again!

My pleasure. :) Let us know how the 2012H's work out for you.

BooBoo Magoo
07-09-2005, 10:55 AM
McDonalds is out.
Walmart's where it's at now Miss Orchid.
They'll be coming to your neighborhood soon.

Alex Lancaster
07-09-2005, 11:28 AM
:p Venerated Miss Orchid:

I think You may find a life companion (none of that marriage stuff) in Ohio, ask for Mr. Giskard.

I think running over students and newspeople with tanks was invented in Mexico City in 1968, there might be earlier cases.

Zilch
07-09-2005, 01:45 PM
Even now we are designing and building your precious JBL speakers. :rotfl:Listen to the Zilchster now, hon:

You'd be better served concentrating on toilets, first.

[And running water to flush 'em.... :p ]

ALSO:

Sex is NOT a function.

[As with ALL matters occidental, seek appropriate guidance from The Rolling Stones in thiis.... :D ]

spkrman57
07-09-2005, 02:32 PM
Someone who no longer posts with us anymore under the same name.

I might be wrong also!!!!!


If it is the same - it is okay with me.


Ron

Miss Orchid
07-09-2005, 03:01 PM
Listen to the Zilchster now, hon:

Please excuse me, zilch is all that I understood from your post.
2012H's working ok now?

John
07-09-2005, 03:10 PM
The part about China taking over the world economy sounds likely if we do not stop shopping at Wallmart. And I think something is fishy in Ohio as well.:hmm: Welcome back:applaud: :applaud: :applaud:

Mr. Widget
07-09-2005, 07:11 PM
Too late.... didn't you see BooBoo's post... Walmart's already off on their next conquest. Soon they'll be touting:

"The Cheapest Junk in the Solar System."

Widget

GordonW
07-09-2005, 07:29 PM
There are JBL systems with Qtc numbers much lower than 0.5 and JBL thought that was beneficial. The readers of this forum are urged to examine group delay figures for sealed systems and vented systems in a variety of box sizes, and in the case of vented systems, a variety of tunings. Sealed boxes can be too small and can decrease power handling capacity due to thermal issues.



There are cases where low Qtc numbers in a sealed box are beneficial. A driver with a pronounced rising response, is not usually one of those situations. It will aggravate the rising response, by "sloping downward" the bottom end response near box resonance.

As for power handling- if that's an issue, then do what JBL did in other cases- use a METAL enclosure, or something similar.

Also, it is a fundamental fact of physics that a damping Q of less than .5, will cause delay, in the sense that it will prevent the initial rise of a signal from reaching full amplitude initially. This will, in any practical context, result in a "dulling" of transients... a "slowing" of the percieved response of the driver.



Also of interest might be the use of sealed boxes as two poles of a high pass filter.


That is exactly what I was implying above. A sealed box rolloff plus a second-order highpass crossover, can be combined to act as a 4th order highpass composite "acoustic filter". If the crossover is set at the box resonance, then the resultant response will have an effective Q of the PRODUCT of the Qs of the two "filters (the box response plus the crossover) multiplied together.

In fact, if the active crossover is state-variable (ie, you can change the Q of the highpass crossover), RAISING the Q of that section a bit, will "EQ" the bottom end of the driver upward, without having to do it in the DSP EQ section. Might result in a lessening of the "digital clip" that was alluded to...

My suggestion for filter tuning, would be to put the speaker in a box that would give a Qtc of between .8 and .9 (.25 to .3 cubic feet), and dial in a Qtc in a second-order highpass filter, of about .9 or so. This would result in a composite 4th order filter with a Q of between .72 and .81 or so. A very subtle boost of the bottom end, and not very much consequence in group delay. Higher slopes should not be necessary, since in the box described, the driver can handle full acoustic power at the box resonance frequency... with a higher-than-second-order highpass composite response, the driver excursion will DECREASE STEADILY below resonance, so there should be no acoustic power handling issues, whatsoever.

Regards,
Gordon.

Miss Orchid
07-09-2005, 08:08 PM
Splendid points GordonW.

There are cases where low Qtc numbers in a sealed box are beneficial.

I do like that statement.

Also, it is a fundamental fact of physics that a damping Q of less than .5, will cause delay, in the sense that it will prevent the initial rise of a signal from reaching full amplitude initially. This will, in any practical context, result in a "dulling" of transients... a "slowing" of the percieved response of the driver.

Yes. There were certain cases where the overall performance desired resulted in Qtc values lower than 0.5 and any dulling of transients was deemed sufficiently inconsequential during extended listening evaluations. The products were put into production and brought to market.

But I digress, this is specifically about the 2012H. Hopefully stevem and Baron030 find all this information helpful. Thank you GordonW. :)

Titanium Dome
07-09-2005, 10:10 PM
:p Venerated Miss Orchid:

I think You may find a life companion (none of that marriage stuff) in Ohio, ask for Mr. Giskard.

I think running over students and newspeople with tanks was invented in Mexico City in 1968, there might be earlier cases.


Check your Czech history of the mid 1950s to see the true enlightenment that tank discipline brings to the unbridled spirit of optimism. Truly crushing the imbalanced liberty of foolish dreamers, the Chi is restored as the negative and positive energy of life is balanced and all aberrations are smoothed out.

Talk about flattening out your response curve: apply this lesson to to your enclosure and even the most novel responder can be brought into perfect balance. :)

Miss Orchid
07-10-2005, 07:38 AM
Baron, I feel like I hijacked your thread! Sorry!

The rest of us can now share in any potential blame? :applaud:

spkrman57
07-10-2005, 08:14 AM
I'll take my share of the blame. This thread has proved to be very interesting.

By the way, did anyone say something about a 2012???


LOL


Ron

Ian Mackenzie
07-10-2005, 04:41 PM
Sounds like the JBL data sheet gives away all the clues.

My Hp Colour Laser printer has product of China stamped on it so the must be getting it right on the score. China has long been the prefered location for multinational manufacturers to globalise. What the lady in the flower pot forgot to mention is that technology and up skilling of the labour comes from outside, so there is a fair amount of control over what's going on over there.

Unfortunately a quick search on another forum reveals illegal breaches of IP from US based manufactures by pirates in China.

The Doctor

edgewound
07-10-2005, 06:37 PM
Well...this has been an interesting read. The usage of certain signature emoticons, as well as linguistics, is quite revealing. Maybe Wal-Mart will own China soon.

geowal3
07-10-2005, 07:17 PM
I have a similar concern about the doghouse for a 2123H. Using BB6, I get results that indicate a similarly sized vented box is somewhat flatter than sealed. This seems "counter-intuitive". What am I doing wrong? Particulars are:

Vented 10 x 16.2 x 6.2, two vents 2.5" D x .75" deep. Fb 89.54 Hz (Shown in red).

Sealed (bb6 suggested) 9.7 x 15.7 x 6 Fb 212.2 (Shown in blue).

Applicaton:A four way system with active 24 db x-over from 150 Hz to 1100 Hz.

Mr. Widget
07-10-2005, 08:21 PM
I happen to be building a version of Daniel's Ancient Audiophile's quest for the Ultimate Home System.


Over the years many people have made reference to a Drew Daniels based or inspired system.... I always shake my head and wonder why. I would hazard to guess that virtually everyone who has made this statement never actually heard the Daniel's System. To pull off a system of that mass and magnitude and not have it sound significantly worse than the sum of it's parts is quite feat. A feat that Daniels was able to accomplish through significant effort aided by his significant skill.

I suppose some people are drawn to it by it's sheer mass.... but today's array PA systems are significantly superior to the Grateful Dead "Wall of Sound" and have a much more cohesive sound. To build a version of the Daniels System and not end up with a smeared wall of sound is almost impossible. I have heard that Daniels himself recommends that people buy LSRs and subwoofers over attempting the "Daniels System."

I am not saying that a scaled down system using some of Daniels' ideas isn't quite possibly a good idea, but to follow it part way without really knowing what you are doing, while quite possibly a real kick in the pants, be prepared to lower your audio expectations.

Widget

speakerdave
07-10-2005, 09:21 PM
This is the big shortcoming in Drew Daniels's system--no response tailoring information whatever. The early versions prescribe a pair of UREI 525 crossovers, but the later system uses some BGW amps with custom contour cards for the various drivers. Go figure. And he does not share the contour information with the reader. I looked at this myself before deciding that it was nothing but a small scale PA, probably a pretty good one, but I doubt it would be at all satisfactory for in-home use, as Widget says. Remember, his target audience is the dedicated crowd that has been listening to D130's with N2400's and 075's since 1957 and now have the leisure to spend some of their children's inheritance on a balls-out system. Drew's to be commended for recommending the LSR32. In fact he was fairly adamant about it; if I find the reference I will post it, but I confirm Widget--it's out there. If you want to make a home-brew large format JBL, study the studio monitor series, including the K2's, and pick one you can get the parts for and copy crossover information. Anything else is fraught with difficulties.

David

Ian Mackenzie
07-10-2005, 09:55 PM
Yes its interesting but these type of driver not intended for anything other than mid range, not mid bass per say, the 2123 is best used sealed in clusters and horn loading.

You would be better served by a 2118H from 150-1100, alternatively a 2220 from 100-300hz then the 2123.

The Doctor


I have a similar concern about the doghouse for a 2123H. Using BB6, I get results that indicate a similarly sized vented box is somewhat flatter than sealed. This seems "counter-intuitive". What am I doing wrong? Particulars are:

Vented 10 x 16.2 x 6.2, two vents 2.5" D x .75" deep. Fb 89.54 Hz (Shown in red).

Sealed (bb6 suggested) 9.7 x 15.7 x 6 Fb 212.2 (Shown in blue).

Applicaton:A four way system with active 24 db x-over from 150 Hz to 1100 Hz.

speakerdave
07-10-2005, 10:29 PM
On the thread below Don McRitchie quotes Drew Daniels in posts #1 and #5 to the effect that he tries to dissuade people from building his high efficiency system in favor of buying LSR32's (and adding a 2245 subwoofer). He is specifically talking about an HT setup, apparently. This, and also the fact that Greg Timbers was quoted as saying something similar were important factors in my deciding to buy a pair of LSR32's on ebay having never heard them. They have been very satisfactory.

David


The thread:

http://audioheritage.csdco.com/vbulletin/showthread.php?t=149&highlight=drew

stevem
07-11-2005, 08:17 AM
Yes its interesting but these type of driver not intended for anything other than mid range, not mid bass per say, the 2123 is best used sealed in clusters and horn loading.

You would be better served by a 2118H from 150-1100, alternatively a 2220 from 100-300hz then the 2123.

The Doctor

Would'nt the same coments apply to the 2012H as well? Isn't it just an updated version of the 2123?

I'm curious to know what others feel is the best approach to a three-way system. Woofer, large format horn and tweeter, or handling the midrange with a cone and then a horn for the highs.

Robh3606
07-11-2005, 08:23 AM
Of the 10" drivers the 2122 has the best lowend performance. The 2123 is smoother above 1.5-2K and a bit more extended as well as significantly more sensitive. These are in sealed enclosures. For me the 2118, 10's/12's cone mids are the way to go with a 1" compression driver.

Rob:)

Robh3606
07-11-2005, 08:24 AM
Just a few ideas from someone who was also inspired by that article



Building the Drew Daniels system Pro’s



It’s Fun

It’s a challenge

You learn a lot

You can end up with a nice sounding system



and Con’s



Lot’s of surprises

It can be a bit of work

You can end up with a not so nice sounding system



It depends what you put into it and what your goals are. If you want audiophile nirvana with imaging so good you can hear the players shift in their chairs stop now. If you want a system that has excellent clarity, terrific dynamics and a vehicle to learn it’s a great experience.



I built my first “version “with 2035’s 2122’s 2416/2344 with subs under and with passive networks based on the article. That changed to an all active set-up with E-145’s 2123 2416/2344 with subs under. If you look at just the drivers it could be called a glorified PA system however it doesn’t sound like one because it’s not just the drivers you have to look and think of the whole system and design it up like that. There is a lot of work like driver layout, tuning’ driver selection, crossover slopes and points and so on. What EQ’s what crossovers, analog or digital, what amps, how much power and on and on. After this I also built a couple of “reference” speakers as comparisons and as references for future projects. Comparing my references XPL-200A and 4344 to my Daniels system a couple of things become apparent. They are all remarkably similar in the sense that they sound like JBL’s if that makes any sense. They all have remarkable clarity and excellent dynamics and are all very well balanced when you put the time in with proper set-up. That said I understand where Drew is coming from. If I had to choose the one with the best overall balance of all as far as imaging’ dynamics, lack of compression it would be the XPL-200 clones. I imagine the LSR’s would be even better. Where the XPL’s get beat is the sense of sheer power the larger systems can project, other than that I think they have the best balance of performance compromises. That said if you feel the urge to build you should. You won’t regret trying to build it however you will regret not trying.



Rob:)

stevem
07-11-2005, 08:33 AM
This is the big shortcoming in Drew Daniels's system--no response tailoring information whatever. The early versions prescribe a pair of UREI 525 crossovers, but the later system uses some BGW amps with custom contour cards for the various drivers. Go figure. And he does not share the contour information with the reader. I looked at this myself before deciding that it was nothing but a small scale PA, probably a pretty good one, but I doubt it would be at all satisfactory for in-home use, as Widget says. Remember, his target audience is the dedicated crowd that has been listening to D130's with N2400's and 075's since 1957 and now have the leisure to spend some of their children's inheritance on a balls-out system. Drew's to be commended for recommending the LSR32. In fact he was fairly adamant about it; if I find the reference I will post it, but I confirm Widget--it's out there. If you want to make a home-brew large format JBL, study the studio monitor series, including the K2's, and pick one you can get the parts for and copy crossover information. Anything else is fraught with difficulties.

David

I must admit that I got into building my own speaker systems as a result of reading Drew Daniels' articles, and in fact, I did borrow some of his design ideas for the speakers I finally built. I read about his systems years before I made my first box primarily because I didn't feel I had the skills necessary to make it sound good. What changed for me was my acquisition of a digital speaker controller. This unit made it possible for me to mix and match speaker elements, crossing them over with virtually any type of slope, and tailoring their responses with a flexible parametric EQ. There is also no need to physically time align the drivers, that function is accomplished by a built in digital delay. I know some people don't care for the sound of these units, but after playing around extensively with high quality film caps and air core coil passive crossovers, I will never go back to them.

Lancer
07-11-2005, 09:15 AM
I'm curious to know what others feel is the best approach to a three-way system. Woofer, large format horn and tweeter, or handling the midrange with a cone and then a horn for the highs.You'll see fantastic solutions later this year or early next year. They won't be cheap and they will be killer.

Just a few ideas from someone who was also inspired by that article



Building the Drew Daniels system Pro’s



It’s Fun

It’s a challenge

You learn a lot

You can end up with a nice sounding system



and Con’s



Lot’s of surprises

It can be a bit of work

You can end up with a not so nice sounding system



It depends what you put into it and what your goals are. If you want audiophile nirvana with imaging so good you can hear the players shift in their chairs stop now. If you want a system that has excellent clarity, terrific dynamics and a vehicle to learn it’s a great experience.Damn straight.

What changed for me was my acquisition of a digital speaker controller. This unit made it possible for me to mix and match speaker elements, crossing them over with virtually any type of slope, and tailoring their responses with a flexible parametric EQ. There is also no need to physically time align the drivers, that function is accomplished by a built in digital delay. I know some people don't care for the sound of these units, but after playing around extensively with high quality film caps and air core coil passive crossovers, I will never go back to them.Nor should you if you are lucky enough to find a solution that negates the use of any speaker level passive components. They really are terrible devices.

On the thread below Don McRitchie quotes Drew Daniels in posts #1 and #5 to the effect that he tries to dissuade people from building his high efficiency system in favor of buying LSR32's (and adding a 2245 subwoofer). He is specifically talking about an HT setup, apparently. This, and also the fact that Greg Timbers was quoted as saying something similar were important factors in my deciding to buy a pair of LSR32's on ebay having never heard them. They have been very satisfactory.

David


The thread:

http://audioheritage.csdco.com/vbulletin/showthread.php?t=149&highlight=drew

If starting from scratch today I would personally opt for the Performance Series instead. LSR6332's would be second choice though, although depending on total cost they might move into top spot.




BTW - pay attention to Drew's warnings. His systems in a typical home have the very real potential to fry your hearing before you know what the hell happened.

Miss Orchid
07-11-2005, 09:41 AM
the lady in the flower pot

Ha! Ha! You very funny!

Mr. Widget
07-11-2005, 10:51 AM
It’s Fun

It’s a challenge

You learn a lot

You can end up with a nice sounding system


ABSOLUTELY! That's why we build DIY projects! I suppose I was just being cautionary... the Daniels System is not really the cookbook design it may seem on the surface.

BTW... speaking of DIY, I think I have found the, "If you want audiophile nirvana with imaging so good you can hear the players shift in their chairs...and a system that has excellent clarity, terrific dynamics..." It has only taken me 30 years and buckets of money!

Widget

Miss Orchid
07-11-2005, 10:54 AM
It has only taken me 30 years and buckets of money!

Patience and persistance.:applaud:

speakerdave
07-11-2005, 12:55 PM
What changed for me was my acquisition of a digital speaker controller. This unit made it possible for me to mix and match speaker elements, crossing them over with virtually any type of slope, and tailoring their responses with a flexible parametric EQ. There is also no need to physically time align the drivers, that function is accomplished by a built in digital delay.


Yes, with digital system controllers, for SR, I believe we've been there for awhile. With upsampling at 24/192 digital processors may have reached a threshold where many more people will find them a better option even for home hi fi. But as Giskard points out, there is the cost. For awhile at least, I think there will still be people willing to struggle with the old technology of passive and active analogue filters, especially if we can keep the network wizards around to help us.

Despite my having the LSR32's to listen to I share your interest in the larger systems. I am, though, a little puzzled by using a 2012 for hi fi. Though out of production it is still possible to put together a 2122, a 10" mid/bass that was purpose designed for monitor applications. If you want a current production model, I think the 2051 neodymium 10", which is used in the 4348, would be worth a look.

David

stevem
07-11-2005, 02:27 PM
Despite my having the LSR32's to listen to I share your interest in the larger systems. I am, though, a little puzzled by using a 2012 for hi fi. Though out of production it is still possible to put together a 2122, a 10" mid/bass that was purpose designed for monitor applications. If you want a current production model, I think the 2051 neodymium 10", which is used in the 4348, would be worth a look.

David

OK, I'm game! Where the heck do I get a couple of 2051s? I thought the 4348 was just available in Japan, but can you buy them (and parts) from JBL in the US?

speakerdave
07-11-2005, 02:53 PM
The 2051 has been on the Transducer list at JBL Pro Customer service for quite a while now. As for the other parts, NOT. Giskard always suggests calling and asking anyway.

(Incidentally, the link to the Transducer List at JBL Pro is not working for me right now; it's crashing my browser.)

David

stevem
07-11-2005, 04:04 PM
The 2051 has been on the Transducer list at JBL Pro Customer service for quite a while now. As for the other parts, NOT. Giskard always suggests calling and asking anyway.

(Incidentally, the link to the Transducer List at JBL Pro is not working for me right now; it's crashing my browser.)

David

Do yo happen to know the part number for the 2051? I checked JBL's transducer parts listing, and I didn't see it. I called JBL parts and they don't list it. I can see them pictured in photos of the 4348, though. Is there another list? Thanks!

Zilch
07-11-2005, 05:15 PM
http://manuals.harman.com/JBL/HOM/Technical%20Sheet/4348%20ts.pdf

Do we mean 2251?

stevem
07-11-2005, 05:55 PM
http://manuals.harman.com/JBL/HOM/Technical%20Sheet/4348%20ts.pdf

Do we mean 2251?

Err, Yep I guess we do! Thanks Zilch!

stevem
07-11-2005, 06:06 PM
I know the woofer from the K2 S9800 is the 1500AL. What is the 1500FE used in the 4348, ferrite? If so, why?

Mr. Widget
07-11-2005, 06:20 PM
What is the 1500FE used in the 4348, ferrite? If so, why?
Yes.

It's less costly.

Widget

speakerdave
07-11-2005, 06:36 PM
Do we mean 2251?

Yeah, it's the 2251J. Sorry about that, sorry to send you on a wild goose chase, stevem, sorry for the confusion, JBL Pro parts people.

David

Zilch
07-11-2005, 06:42 PM
Yeah, we told them we wanted what Speakerdave said.

[And, NOTHING else would do, thank you.... :p ]

speakerdave
07-11-2005, 07:54 PM
Yeah, we told them we wanted what Speakerdave said.
That should have done it for you. Try again tomorrow.

David

stevem
07-11-2005, 08:01 PM
Yeah, it's the 2251J. Sorry about that, sorry to send you on a wild goose chase, stevem, sorry for the confusion, JBL Pro parts people.

David

No problem. With this hobbie, I'm used to chasing wild geese! Oh, and I told the parts person at JBL that I might not have the right driver number, so we didn't spend a lot of time on it.

stevem
07-12-2005, 11:21 AM
Despite my having the LSR32's to listen to I share your interest in the larger systems. I am, though, a little puzzled by using a 2012 for hi fi. Though out of production it is still possible to put together a 2122, a 10" mid/bass that was purpose designed for monitor applications. If you want a current production model, I think the 2051 neodymium 10", which is used in the 4348, would be worth a look.

David

I did a quick plot of the 2251J against the 2123H, and they look like virtually the same driver. I would think that they would each be equally suitable for use in a studio monitor. What do you think?

speakerdave
07-12-2005, 11:42 AM
Don't know. Are you talking about a response graph based on T/S parameters plugged into a box program? There might be differences between these two drivers that are not shown by that. My whole point was that I know so little about this stuff that in making plans for speakers I usually stick to applying JBL drivers the way JBL used them. One reason is that I might be able to use existing crossover and filtering engineering. But other forumites are more intrepid than I am. I know that at least one of our members has used the 2123 and one has used the 2012 in a four-way monitor type speaker and has expressed satisfaction, so they may be a source of information about using those drivers that way.

The idea of using the 2251 is just one idea I've toyed with in thinking about a four-way monitor. Right now I'm off that track entirely and am barking up another tree. The crossover information for the 4348 is available, incidentally.

I'll see if I can find those posts.

David

Baron030
07-12-2005, 11:00 PM
Hi Everyone
I just want to say thank you for responding to my questions about the 2012h driver. I would have gotten back sooner, but I have been busy wiring up my last two amplifiers. All of you have given me a lot to think about and a quite few things to try out as well.
I will try to keep you posted on my progress.

The project that I am working on is extremely challenging and very expensive to complete. It sure would have be a lot easier and cheaper to go out and buy a great speaker system right off the shelf. But, sometime the journey is just as important the destination. And I am having a lot of fun building my own system from scratch.

Well, here is a listing of the components that I have already purchased and a rough idea of how I plan to put them to use. Starting with the bass, I have purchased a pair 2226H drivers and plan on mounting them in 5 cu ft. enclosures tuned to 40 Hz. For low mid-range, I am going with the 2012H drivers. And at this time, I have not decided on the enclosure design for the 2012s yet. For high mid-range, I have a pair of 2446H compression drivers with 2382A horns. And finally, for the high end, I have a pair of 2405 tweeters. As for the crossover network, I chose an Ashly XR-4001 stereo 4-way electronic crossover. It was reasonably priced and it should provide me with a lot of adjustments for balancing out the system as a whole.Currently, I am thinking of using 400Hz, 1000Hz, and 7000Hz as the cross over points.

In keeping, with Drew Daniel’s recommendations for the ultimate home systems, I am going to be driving my speakers with insane amount of reserve power. I have decided to go with a strange mix of “high tech” and “old school” for amplifiers. For the bass frequencies, I have purchased a Crown K2, which can put out 500 watts per channel. For low mid-range, I have purchased a Crown K1, which can put out 350 watts per channel. Since, the compression drivers are so much more efficient, I have opted for some “old school” Crown D series amplifiers. A pair of D75A Amplifiers wired in bridge-mono power the 2446H drivers with up to 110 watts per channel. And finally a single D75A Amplifiers powers the 2405 tweeters with up to 40 watts per channel. As a whole, that works out to insane 1000 watts per channel. And here another real surprise, there are no cooling fans. Why, without fans, this system is completely silent, until… http://audioheritage.org/vbulletin/images/smilies/rotfl.gif

Baron030

speakerdave
07-13-2005, 01:08 AM
stevem:

The 2123 was used in the 4344 Mark II. Haven't found any of the details yet.

David

stevem
07-13-2005, 10:41 AM
stevem:

The 2123 was used in the 4344 Mark II. Haven't found any of the details yet.

David

Thanks, Dave. At least JBL used it in a monitor.

Baron:

This is very preliminary, but at this point I would recommend a small .3 to .5 cubic foot sealed enclosure for your 2012H mids. I built a pair of ported boxes, and they don't help to raise the falling response I get at 300 Hz like I had hoped. They do bump up the response lower down at about 125Hz, but this is far below the crossover frequency and has little effect. The over-all sound is a little warmer and laid back with the ported boxes (I really don't know why, and it's very subtle), and a little more up-front sounding when I close off the ports. It sounds good either way, and I'm still fooling around with it. I'll let you know if I decide on one way or the other. Good luck on your project.

Robh3606
07-13-2005, 11:40 AM
Hello Baron 030

"In keeping, with Drew Daniel’s recommendations for the ultimate home systems, I am going to be driving my speakers with insane amount of reserve power."

If you don't have an SPL meter you should purchase one. With my active set-up I am under powered for my mids and midwoofers. I have 100 watts per driver. I have never been able to get the Crown IOC lights to even flash on the PS-200's even at stupid your loosing your hearing levels. I use the 2123 as my midrange and it sounds just fine.


Rob:)

Baron030
07-18-2005, 11:05 PM
Having purchased a pair of the 2012H transducers, I quickly found that without some form of equalization. The rising response of these drivers proved to be completely unacceptable.

Since, I was planning on using the 2012 transducers as part of a 15” - 4 way quad amplified system. I knew that I could insert a passive equalization network somewhere between the crossover and the 2012 driver and not have to worry about how the network would interact with the other drivers in the system.

So, I set about examining the JBL published response chart under a 10x magnifying glass and transcribing 130 data points, into a data base file. Since, I am a computer programmer by profession. I was able to write a few programs that could test several different types of passive filters. After testing a few dozen possible designs,, I found that by simply wiring a 1.5mH coil in series with the 2012 driver. It appeared to have flattened the response over a fairly wide range. Since, the DC resistance of the coil can reduce the damping on the driver. I would recommend using an air core inductor with as low DCR as you afford.

And I was going to include an excel chart as part of this response. But, I am not sure how to go about doing that. Just because I’m a programmer, does not mean that I know very thing about internet.

So, I will have to summarize the important points of the chart. The 1.5mH value was chosen to optimize for the 400Hz to 1000Hz range of my own project. And coil caused the sensitivity to rise by only 0.06 Db between these two values. But, unfortunately the response is not perfectly smooth. So, response has a 0.29 db peak at 420Hz and a nasty -1.66 dip at 740Hz. and below 400Hz, the response gently rolls off -1.60Db at 230Hz. And above 1000Hz, the response gently rolls off -1.66Db at 1475 before dipping even further. And the two big peaks at 3166Hz and 4700Hz are both suppressed to -0.56Db and -1.32Db respectfully. Please realize that these response numbers are just a computer simulation based on the published response chart and may not be this accurate.

But, I have recently made some listening tests. And the 1.50mH coil is making 2012H drivers sound really nice. And all of the harsh peaks are now gone.
Hopefully in the next few weeks, I will be really putting this combination under some serious tests. I have purchased a DBX RTA-M reference microphone and with the use of a musician friend’s end high end microphone preamplifier. The combination should give me an accurate sound pressure meter for measuring the JBL2012H+1.50mH response.

Baron030

Lancer
07-19-2005, 07:27 AM
And I was going to include an excel chart as part of this response. But, I am not sure how to go about doing that.I usually copy the Excel chart into Paintshop Pro and then resize it to 1024 width before posting it as a jpg.

Let's see what happens with MS Paint.
Open Excel document.
Select chart.
Ctrl-C
Start -> Run... -> mspaint
Ctrl-V
File -> Save As
filename is untitled and save as type is jpg
Manage Attachements
Browse and Upload

Lancer
07-19-2005, 07:35 AM
2012H in 0.5 cu ft sealed box.

Lancer
07-19-2005, 07:37 AM
Add in the Parts Express 1.5 mH 14 AWG air core

tomp787
07-19-2005, 03:26 PM
Hello,

There are cases where the DCR of the series inductor can be used to your advantage. It effectively raises the Qes of the driver thereby changing the low frequency rolloff. Here is the equation:

Qes 2 = [ Qes 1 x (Re + Zo)] / Re

Where:
Qes 1 is the driver's original Qes
Qes 2 is the new Qes
Re is the driver's VC resistance
Zo is the amplifer's output impedance (or series R)

Tell us what happens when you model some inductors with higher DCR. My guess is a higher DCR (cheaper?) inductor will bring up the low end response but maybe the dcr is insignificant? In either case you may win.

Regards,
Tom

Baron030
07-19-2005, 08:45 PM
Thanks, Lancer
Call it a brain fart, But, I just could not figure out to how to attach the chart last night. Somehow, I just knew it was going to be easy, once you can get a handle on it.
As I stated before, I chose the 1.50mH value because it appeared to give me the flattest response over the 400Hz to 1,000Hz range for my project. And that I will be using this inductor with an electronic crossover network. For those who are planning on using a passive crossover network, the 1.50mH coil will not work. Instead, you will need to design a network with a dropping response to compensate for the 2012H drivers rising response. Since, a 1.5mH coils is equal to a -6Db per octave low pass filter with a 850Hz crossover point. That should give you a clue as to where to start on a passive crossover network design.

http://audioheritage.org/vbulletin/attachment.php?attachmentid=9066&stc=1

stevem
07-19-2005, 08:48 PM
Strictly from a subjective point of view, I'm liking the ported enclosure a bit better than the sealed one. It sounds a little more "relaxed", although it appears about the same within it's pass band (with my admittedly coarse measurements).

I also find it interesting that in a Stereophile review of the Wilson Maxx something, they are also porting the midrange enclosure, and the port doesn't contribute much output in their design either. I'm going to give the 1.5 mH inductor a try too, I must have a couple of that value around here somewhere.:hmm:

stevem
07-19-2005, 09:15 PM
By the way, I agree with you Baron that the rising response of this driver is unacceptable. From the begining I have used active EQ to flatten the rise, and also a boost at low end to flatten the response to just below my crossover frequency of 300 Hz. I found that this "splitting" the EQ instead of starting at a lower frequency and using a lot of cut only (but with global gain to bring the level back up) seems to sound a little better in my system.

GordonW
07-20-2005, 09:50 AM
This "splitting" of the EQ, ie, slightly boosting the lows and cutting the highs, does make sense, from two standpoints:

1) The EQ is probably cleaner, nearer the "center" of its range... this is frequently the case with these types of devices. EQs tend to behave quite a bit better, when they're not asked to make extreme changes to a signal.

2) The phasing/damping is probably improved, overall. Boosting the bottom end will effectively reduce some of the damping, in what is almost always an over-damped bottom end response with these drivers (unless they're in a TINY enclosure). This will restore some of the transient rise behaviour (ie, improve the time response of the driver to midbass transients)... ie, the driver will react "faster" to transients in the 300 Hz range or so, bringing its time-domain behaviour in that range to something nearer its time-domain response at other frequencies above that.

Regards,
Gordon.

Rudy Kleimann
08-14-2005, 09:52 PM
Something else that factors in to this comment: Phase shift by the EQ itself. All analog Equalizers introduce phase shift as their filters are activated to introduce an output level change within their passband. the more EQ, the more phase shift.

Not to mention the ragged frequency response curves between the center frequency of each EQ filter when large EQ adjustments are made yields a FR curve that looks more like a rollercoaster track when running through several adjacents bands on the EQ.

Always try to spread your EQ'ing equally above and below the mean (Zero) level, i.e. start a little above 0, and end up a little below 0 on the other end of the long slope you are trying to make with it.

Baron030
09-27-2005, 03:31 PM
Well, I have been very busy for the last few months. But, I have finally had the time to do some testing of the 2012h drivers. The charts below were created by using the following equipment and procedures. A DBX RTA-M microphone was used with a PreSonus tube preamp and a Rives Audio Test CD. Unfortunately, the Rives CD contains sine wave test tones spaced 1/3 octave apart and not 1/3 octave filtered pink noise. Sine wave test tones are very sensitive to room resonances, while 1/3 octave filtered pink noise will tend to average out some of a room’s resonances. As a work around for this problem, I made a dozen measures varying the placement of the drive in the room and varying the distances between the diver and the microphone. In the future, I am going to track down a CD of 1/3 octave pink noise and save myself a lot of time. What really surprised me in testing was seeing a roll off above 1000 Hz. The published response curve of this drive clearing shows a rising response above 1000 Hz. As you can see, the drive does have a nasty peek around 800 to 1000 Hz. Without any equalization, this peak can easily be heard during listening tests. To provide some equalization, I have experimented with using a 1.50mH coil. As you can see, as a very simple equalizer, the 1.50mH coil works fairly well. But, I do see the need for adding a resister and maybe a capacitor as well, wired in parallel with the coil to really smooth out this driver’s response.
http://audioheritage.org/vbulletin/attachment.php?attachmentid=10246&stc=1

http://audioheritage.org/vbulletin/attachment.php?attachmentid=10247&stc=1
Baron030

duaneage
09-27-2005, 09:34 PM
China seems to have great faith in the US, that is why they are sitting on over 200 billion doillars in US currency. Like it says " In God We Trust"


Chinese money is called Remimbe, loosely translated as The Peoples Money. If only that were true.
I never met anyone in my travels there that did not want to live in America. That is why the Chinese name for the US is Mei Gwok.

And that means "Beautiful Land"