Jbl 1200fe-8

[Dr.db]

I just bought a couple of these 12" woofers which I am going to use in some bookshelf speakers.

I want to follow an existing JBL enclosure-design so I´m going to build some test-boxes first. So far I´m aware of the 4428, 4429 with aprox. 1,8 cubic ft3 and the S5800 with aprox. 2,25 cubic ft3.
Is there another system using the 1200FE with a different enclosure?
The published specs from JBL actually do confuse me a little. The smaller 4428 is specd with -6db at 40hz while the bigger S5800 is specd with -6db at 50hz...

What are your thoughts on these enclosures?
Is it worth building an even larger enclosure than 2,25 cubic ft3, or is the 1200FE specifically designed for small enclosures?

Best regards,
Olaf

[RMC]

Hi Olaf,

I don't think the 1200fe was made for use with larger box volumes. With more volume than you mentioned the driver shows signs of being stretched. I tried 3 cu.ft, Fb 37hz, F3 36 hz, to try to make it flat and overstretching is evident between the low to mid bass range with a quite noticeable drop of 1.3 db or so. I think its designed for smaller volumes in mind. I used Winspeakerz to model these, Ql 7 as usual.

I've also modeled 2.25 cu.ft, Fb 40 hz, F3 is at 38 hz, response shows some drop between the low to mid bass region.

I've done 1.8 cu.ft., Fb 40 hz, F3 is at 42 hz, response drop of 0.5 db below 125 hz. Using same volume but Fb 42 hz leads to F3 42 hz but response is flatter within a small fraction of a db.

So with my three quick modelings the best bet appears to be the last one considering the only slight differences in F3 and the flatter response curve obtained: 1.8 cu.ft, Fb 42 hz, F3 42 hz and response remaining within about a third of a db. Regards,

Richard

[Dr.db]

Hi Richard,

thanks a lot for your effort.
I had simulated the FE1200 with WinISD too and came to the same conclusion, but I have to do some real life testing to prove it next.

Do you know of any JBL-system containing the FE1200 using a larger enclosure than the S5800 ?

[grumpy]

Maybe nicely ask TiDome to measure the boxes Greg had built.
They sound quite nice and are real-life examples.

[RMC]

Hi Olaf,

I don't know if there is any JBL system using a larger enclosure than those you mentioned for the 1200FE. My impression is I doubt very much there would be one. Simply because of the modeling results I got in Winspeakerz showing driver being stretched as box volume increases above the volumes you gave (quite noticeable response drop between the low and mid bass range). Typical response of a drivers made for small volumes when used in larger box. One has to tune the box higher to get somewhat acceptable response, however the low to mid bass response drop remains...

My 12" 2214H are better behaved on this aspect since they have almost double the Vas number of the 1200FE (224 L. VS 125 L.). The box size ratio (alpha) is Vas/Vb.

Seems Bill Mc Fadden has provided one of the best simplified explanations I have yet seen in relation to Fs, Qts, Vas:

"2. Small Signal Parameters

The three parameters that primarily determine the frequency response of a loudspeaker are compliance, free-air resonance, and Q.

The compliance, Vas, is a measure of the overall stiffness of the cone, surround (the part the attaches to front of the cone), and spider (the part that attaches to the rear of the cone). It is specified as the volume of air having the same compliance as the driver. A small number corresponds to a small volume of air, which is stiffer than a larger volume of air. Thus, compliance and stiffness are inversely proportional. Optimum enclosure volume is proportional to Vas.

Free-air resonance, Fs, is the resonant frequency of the driver's voice coil impedance with the driver suspended in free air (no enclosure). The -3 dB frequency (F3) of an enclosure is proportional to Fs.

The Q, Qts, is a measure of the sharpness of the driver's free-air resonance. It is defined as (Fh-Fl)/Fs, where Fh and Fl are the upper and lower -3 dB points of the driver's voice coil impedance in free air. Optimum enclosure volume is related to Qts but is not directly proportional. It is accurate to say that the volume gets larger as Qts gets larger. Likewise, F3 gets smaller as Qts gets larger, and for the sealed box enclosure, F3 is inversely proportional to Qts." (RMC: and optimum volume gets smaller as Qts gets smaller, other things being comparable). (Bill Mc Fadden, Loudspeaker Primer, From his Web page, rdrop.com/users/billmc/).

Your driver's Qts is pretty normal at 0.26 so this one doesn't seem to be the issue. Instead, it looks more like the rather low Vas number of 125 L. is responsible. And since "Optimum enclosure volume is proportional to Vas", well low Vas number means smaller optimum box size here. The woofer's Design Engineer (Jerry Moro) may well have chosen a stiffer suspension system for the purpose of good performance in smaller box volumes.

BTW, I have not seen in Moro's 6/11/03 specifications data a number for Xmax which surprised me. Maybe I'm blind. In any case when modeling the driver with Winspeakerz the software assigned a 6mm Xmax number, calculated number or ballpark figure?, since I didn't see a number to enter from Moro's specs. 6 mm MAY be a reasonable number here VS driver's stiffer suspension. The 2214H is at 6.6 mm but its a much older driver than 1200FE and the latter may have more than 6 mm...

With regards to test boxes: remember, test boxes have to be identical in quality to the real thing (except for finish) to get meaningful results otherwise its a waste of time to make a quick, cheap box so so rigid, braced and sealed, and to expect truthful or representative results from a "donkey" VS what the originals should/will be... Instead of doing the work twice, I prefer to save time and material by building only one test box, very well made, test it and if its OK you have only one left to do. If not OK then you have only one box to scrap or reuse on something else...

Since you mentioned using Win ISD don't forget to set each time the QL number to the industry standard 7 instead of the default 10 used in Win ISD Pro to get more acceptable and comparable box results with other softwares. I mention this simply because many people forget to change the Win ISD Pro QL number each time they start a new modeling. Sometime ago a fellow member here asked me what software I used to model his driver/box since his results were different than mine. He forgot or didn't know he was using QL 10 instead of the standard QL 7. I do have and use Win ISD Pro 2016 edition also, but Winspeakerz is faster, more flexible, low cost and has a much larger driver database (1,500) that I keep adding to VS Win ISD Pro very limited database. Using QL 10 as a starting point is overly optimistic for most box designs. Using 7 instead of 10 leads to a little larger box to compensate for losses (e.g. air leaks). Regards,

Richard

[Dr.db]

@Richard:
Thanks a lot for this detailed answer, I got a deeper understanding now.
Actually I made the mistake you mentioned with WinISD, thanks for that!
It absolutly makes sense to me the 1200FE is designed for small enclosures, so I will do the same. As these are supposed to become bookshelf speakers for my small room, I actually like the idea of a small enclosure a lot!

I couldn´t find any info´s about the linear xmax as well, propably because these drivers where never meant to be sold seperatly...


@Grumpy:
Thanks for the hint.
But if I remember correctly, Greg Timbers used these with subs below. But in my case I will be using these as the low-mid woofers without sub, so I need full extension....

[grumpy]

You remember correctly. :)
However, the subs simply augmented what is already an excellent bottom end (having heard them only without the subs, I did not feel there was a missing bottom).
I'm just thinking the measurements might be another useful data point in your assessment. I expect you will enjoy the result, however you arrive!

Best regards.


@Grumpy:
Thanks for the hint.
But if I remember correctly, Greg Timbers used these with subs below. But in my case I will be using these as the low-mid woofers without sub, so I need full extension....

[pos]

Maybe nicely ask TiDome to measure the boxes Greg had built.
They sound quite nice and are real-life examples.
Don't know if these are real clones of GT's arrays, but should be quite close:
http://www.audioheritage.org/vbulletin/showthread.php?23458-Two-Project-Array-1200-projects&p=242312&viewfull=1#post242312

70.8L tuned to 32Hz

[RMC]

Hi Olaf and Pos,

Attached is a picture (sorry for quality) of the modeled response I get with 1200FE/70.8 L or 2.5 cu.ft./Fb 32 hz. Done with Winspeakerz, half-space loading, QL 7. The red dotted line in the graph represents the 6 mm Xmax line done by the software as I didn't input that number. The red solid line at the bottom of graph is excursion at 1 watt.

LF response looks pretty droppy to me starting at about 300 hz and down to - 2.5+ db at 40 hz. Appears like a too large box and/or too low tuning curve to me. Doesn't seem very appealing for a bookshelf stand alone unit. Will sound bass shy vs the rest of the spectrum? Regards,

Richard



82859

[grumpy]

I’m all for educated simulation and device measurement before cutting wood and efforts to understand the difference between a simulation and what is ultimately measured ... and what is heard.

What I can say is the integrated system was not lacking in bass
... at least not to my taste :) I have no in-room measurements to back it up (maybe next visit :) )

My recollection is that the same was said about the system while at Greg’s place with subs off.

In both situations, the system was not against a wall but somewhat into the room and on stands (roughly the same position from the wall as a later GT system I -did- have the pleasure of hearing in his LR)
Closer to a wall would put the system nearer to a 1/4 space environment. In any case, both fairly normal listening environments hosting the 1200 driver in Array type systems sounded quite nice. Which was really all I wanted to say. Other design goals and preferences will lead to other places/choices. Viva la difference.

[RMC]

Hi Grumpy,

Was there any form of signal processing involved in the listening test, EQ or other?

I'll tell you why the question after. Regards,

Richard

[pos]

It depends on how you look at it: this dropping response down low can be seen as a rising response in the higher range, and addressed in the electrical LP filter to get the proper acoustical response with a lower final sensitivity.
In an active/digital system you would typically use a shelving filter, and in a passive system you would lower one or two poles of the filter.

I think GT posted the voltage drives of his project array 1200 somewhere, but I cannot find it right now.

[grumpy]

Hi Grumpy,

Was there any form of signal processing involved in the listening test, EQ or other?

I'll tell you why the question after. Regards,

Richard

My recollection is that only the passive crossover was at work. No active signal shaping or tone controls, DSP, electronic delay, etc... between source/pre/amp/speaker.
I'll also look for the crossover schematic (I -should- have it somewhere ... perhaps in a digital photo which I have -far- too many of ... will have to look at some offline storage :o:)

[pos]

schematic:
82868

acoustical response: (LR 24dB/oct @ 700Hz crossover)
82870

[pos]

voltage drive for active biamping, probably quite close to the passive one for the low-pass part:
82871

[pos]

I traced the active low pass curve and got pretty close using a combination of:

2nd order low pass filter at 240Hz, Q=0.68
2nd order low pass filter at 650Hz, Q=1.58

[RMC]

Hi Grumpy and Pos,

Re my "Will sound bass shy vs the rest of the spectrum?"

@ Grumpy:

Thanks for the info on the 1200FE listening test. Its informative.

After making the 1200 modeling in Winspeakerz, I did another quick one this time with Win ISD Pro thinking my first one may be incorrect (wrong entry from me or something). But results were the same the 2nd time.

"My recollection is that only the passive crossover was at work. No active signal shaping or tone controls, DSP, electronic delay, etc"

I just wanted to rule out this possibility of signal processing impact. Your listening test is in not being challenged here.

In fact I'm testing the following statement, made in the help files, by the makers of Win ISD Pro with regards to the frequency response curve generated by their software, i.e. the green plot they refer to below:

"The green plot is the actual representation of how this speaker will sound in a completed box." They assume the usual half-space driver radiation as they mention.

The listening test doesn't support their statement (according to which it should have sounded bass shy but it did not), moreover they seem to forget other important aspects of room acoustics effect on speakers with such statement... Therefore, Grumpy 1, Win ISD 0.

@ Pos:

RE: "It depends on how you look at it: this dropping response down low can be seen as a rising response in the higher range, and addressed in the electrical LP filter to get the proper acoustical response with a lower final sensitivity."

You're absolutely right, I agree with you. I keep forgetting about this. Similar to the glass that can be seen as half-full or half-empty, depending on how you look at it... Regards,

Richard

[Robh3606]

I couldn´t find any info´s about the linear xmax as well, propably because these drivers where never meant to be sold seperatly...


You have a 1" coil with a .5" gap so figure about 12.5mm PP maybe a little less.

Rob:)

[Dr.db]

Thanks a lot for all this great input and support!

My assumption is, that JBL used a small enclosure (1,8 cu ft) within the 4428 modell to justify it as a bookshelf speaker. With the S5800 being a big floorstand-speaker, they aimed for a more appropriate enclosure volume... Just my guess...

@Rob: That sounds logical and is much better than the Audax-woofer which I was going to use in the first place.

[JuniorJBL]

I would add to Grumpy's statement of the sound of these 12" 2 ways. It was probably the best sounding 12" 2-way I have ever heard. We both listened to them in the same spaces two different locations. :applaud:

[Ian Mackenzie]

I agree with Grumpy’s comments.

If you look at the voltage drive on the woofer the signal to the woofer is shaved off or contoured above 100 hertz.
It’s a passive EQ if you like.

The woofer tuning is what is referred to as a shelf alignment which is a tuning used in LEAP software.

The actual acoustic response sums flat when the HF level adjusted to match the EQ’d response of the woofer.

The loss of 2 dB or so mid band sensitivity is an acceptable compromise.

Jbl do do this elsewhere in other systems and you can see it in the voltage drive on the woofers.

As Richard points out it would otherwise be unbalanced subjectively.

[RMC]

RE post # 21

Certainly the best post here to align logically the different views mentioned (not necessarily contradictory ones) about sound heard (Grumpy and Junior), modelings done (me), crossover details and way of looking at response curve (Pos), etc. That reconciliation of the info does make sense to me! Not afraid to say it, a good shot is a good shot.

Since I didn't hear/don't own the Array, 1200FE nor LEAP I can't really comment on actual sound heard for example.

GT and many others at JBL are quite smart Engineers. They can make almost any speaker system and still make it sound proper, with the number of "tricks" they have up their sleeve.

The alignment chosen is not the first go to or must have type of LF alignment I would jump on normally for a few reasons, but the Design Engineer has his reasons, tricks and constraints to go that way, which I respect, and these are certainly as good or better than mine.

It seems to me one of the objectives may be to extract some deeper bass than usual from a relatively small box Vb, even if that means losing a couple of DBs on sensitivity. Its a purposely made design choice, as others are.

Richard

[Dr.db]

Indeed, very well thought answers! :)

If I´m not mistaken, this passive-eq or high-shelving of the woofer occurs in the 4428/4429 circuit as well. There are two inductors in series with the woofer...
My plan was to use these with a JBL M553, so I´m stuck with a simple 4th order filter. I guess I should rather aim for a more conventional (linear) tuning than. This means smaller box with higher tuning...

[Ian Mackenzie]

What l would do is acquire a dsp amp with some EQ and level the shelved bass alignment with a simple shelf filter below 100 hertz with +2-3 dB lift.

If your signal path is pure digital you can do this in a variety of ways.

[RMC]

Hi Olaf,

In case you don't want to add another amp or if DSP ones are too expensive, you may want to consider another option, Parametric EQ boost to deal with the shelf issue. Though the example shown below is a single channel unit, XLR, 1/4" and RCA, two of them cost $320.US ($160. each) and the Euro is worth about 12% more than USD... Made in the USA shown! I don't know if they are available directly in Germany though, you would need to verify that. Regards,

Richard


82922

[pos]

The required LP filter is easy to achieve with an analog crossover using Sallen-Key topology: you simply need to chain two LP sections with the proper freq and Qs measured in post #16.

[Dr.db]

What l would do is acquire a dsp amp with some EQ and level the shelved bass alignment with a simple shelf filter below 100 hertz with +2-3 dB lift.


Guess what, my NAD C165 Preamp has a 100hz shelving-filter (Tone-controls...). :)

@RMC: PEQ are great, I can borrow a Klark Teknik DN405 to try...

@Pos: It might be easy, but I want to keep the JBL M553 untouched.

[RMC]

Hi Olaf,

The important part I'm glad about is the outcome here. At least now you have different options to consider/try: tone controls, DSP amp, Parametric EQ, custom built filter (Pos). Having worthwhile options to choose from, based on your specifics, is a lot better than having none and your back to the wall... Regards,

Richard

[Dr.db]

Well summarized Richard, now it´s up to me ;)

I will have an Audax PR330M0 12"-woofer for comparison, but I suppose the JBL will produce a better low-end.

[RMC]

Hi Olaf,

RE: "I will have an Audax PR330M0 12"-woofer for comparison, but I suppose the JBL will produce a better low-end."

Your Audax PR330MO remind me I did use a pair of Audax 10" woofers in boxes years ago, but can't remember the model number, it may have been an OEM model. Unfortunately, at one point the foam surrounds were really done and I couldn't find the right replacement for these. The foam roll on that model was notably wider than usual for driver size, and I didn't want to spend time and money on making Frankenstein woofers with these... So I gave them away.

I kept the good boxes (Vb about 1.5 cu.ft.) and with some minor modifications I made a pair of small sound reinforcement boxes using Britain's Fane Pro 10" woofers with special cone and Vifa dome/horn tweeters. I bi-amp these enclosures with the 2205H bass cabinets I have (not subwoofers). Nice SR Sound. I've been using some European drivers from various manufacturers (Fane, Vifa, Peerless, Audax) for a while, most of which I still have.

As for your JBL/Audax comparison, on paper, I note the following: Fs 27/28 hz, Qts .26/.27 so these two items are pretty much the same; Vas 125/264 L. The Audax has more than double the JBL here. Based on the principle that optimum box volume is proportional to Vas (Bill Mc Fadden) there's a possibility, depending on box size used, that the box may be a little small for Audax optimum LF potential, therefore creating a small LF bump, though this may not be a big deal in practice.

With regards to Xmax 6.25/4 mm the JBL is a little more generous on this aspect for low-bass reproduction, as well as on power input capability, however on sensitivity 91/98 db if real the Audax would be far more sensitive which can make-up for its lesser power capacity of 150W. I say "if real" because one graph seen shows closer to 94 db and the other graph seen is closer to 98 db. I tend to think the former number makes more sense than the latter for a few reasons, but I may be wrong.

In any case, when comparing sounds with same input power, don't forget that a more sensitive and louder speaker usually seems more impressive, but its not necessarily better sounding than a less sensitive one. You'll decide which one pleases you most. Regards,

Richard

[Ruediger]

Maybe I am blind: I can't find Vas for the 1200fe

Ruediger

[RMC]

Hi Ruediger,

Its in the "Engineering test and performance specifications" of the Design Engineer Jerry Moro posted on LH July 23, 2008, revision C dated 6/11/03. Go down to the MLSSA Parameters showed, line 12: 125 L. Sorry, I'm not good at all for posting links... Regards,

Richard

[RMC]

Hi Ruediger,

What about this. Good enough? Regards,

Richard





82958

[Ruediger]

QB3 Thiele alignment #3 would match if Qt remains unchanged (no passive crossover), and
QB3 Thiele alignment #4 would match if Qt gets changed (passive crossover).
Taking these two alignments as starting points other alignments could be found by interpolation.

Ruediger

[hlaari]

JBL 1200Fe thiele/small

http://petoindominique.fr/php/mysql_thiele_seul.php?hp=4333

x-max is 5mm one way
x-max pp is 10mm




Ari

[RMC]

Hi Ari,

These Xmax numbers are a little worst than those posted here previously ("12.5 mm PP" or 6.25 mm one way), or calculated or assigned (6 mm) in my Winspeakerz software. With regards to the 5mm number Petoin indicates that its the "data base value". Which data base is he referring to?? I haven't seen any other than Jerry Moro's (Design Engineer) numbers where there's no specifically mentioned Xmax number.

Personnaly, I would have hoped seeing an increase in these numbers, not a drop, considering the low box tuning mentioned previously.

The woofer is described as High-power (the power test shown in the engineering specs gives a good indication of this), then the box is low tuned (putting more strain on a driver) and the 12" speaker has somewhat limited excursion capability, even more so in Petoin's data... Maybe the Engineer had another trick hidden up sleeve? Otherwise something would have to give at one point. e.g. Power input? Distortion? Regards,

Richard

[hlaari]

Hi Ari,

These Xmax numbers are a little worst than those posted here previously ("12.5 mm PP" or 6.25 mm one way), or calculated or assigned (6 mm) in my Winspeakerz software. With regards to the 5mm number Petoin indicates that its the "data base value". Which data base is he referring to?? I haven't seen any other than Jerry Moro's (Design Engineer) numbers where there's no specifically mentioned Xmax number.

Personnaly, I would have hoped seeing an increase in these numbers, not a drop, considering the low box tuning mentioned previously.

The woofer is described as High-power (the power test shown in the engineering specs gives a good indication of this), then the box is low tuned (putting more strain on a driver) and the 12" speaker has somewhat limited excursion capability, even more so in Petoin's data... Maybe the Engineer had another trick hidden up sleeve? Otherwise something would have to give at one point. e.g. Power input? Distortion? Regards,

Richard

You might be right, but I think at the best way is to get someone to measure the thiele/small for 1200Fe to get everything 100% right
I own pair of 1200Fe and they one of the best 12" woofer out there

I am now working with the big brother of 1200Fe, the 1500Fe
I am getting close to have all parts I need to make pair of 4348 monitor, this project is a challenge, and I hope at I can get the wood work and veneering as perfect as I can


Ari

[Dr.db]

@Richard: Yes, I will have to check the surrounds and spider on these Audax PR330M0, I don´t know how old these are untill I see them.
I have been told they sound really nice, but are missing the last octave down low.

@Rüdiger & Ari: Thanks for your additional infos!