1400Nd Vs 1200Fe (2- way with 1.5 horn)

Hi Didier,

I'm somewhat surprised that you seem to feel obliged to use a 120 L. box so that you don't regret later, whereas I've given you in post # 12 a few smaller box options that would be better for your living room space as you say (e.g. Win ISD Pro (2016) QL 7: 85 L. (3 cu. ft.), tuned at 40 hz, also pretty flat response except for a small .75 db drop around 85 hz, F3 @ 40 hz). Rob has added the suggestion of low frequency DSP to increase the bass.

My word of caution on that was "... isn't a good idea in my view even if you compensate with DSP (i.e. equalization) because it requires more power which leads to more distortion caused by increased cone travel..." In a home hi-fi setup, as Rob mentions, where one doesn't usually try to rattle the planet, if you have sufficient power available to provide + 6 db LF boost (= 4 times the power) and the driver is capable of reasonably high excursion (1400 ND: Xmax 7.62 mm, good) then why not go for it ? I would, with or without EQ.

I mentioned in post # 20 "A steep high-pass filter is usually required as E-V doest it with their "step-down mode EQ" and there are other conditions like, if my memory is correct, tuning half an octave below "normal", not just any frequency below... You should read more on this issue." My point here was that you don't pick an Fb at 30 hz because Rob did so. It all depends on each specific case. The above-mentioned box of 85 L. with a proposed tuning at 40 hz would lead to a 30 hz tuning if designed for "Step-Down Operation". Another box another tuning...

I've pulled-out my tech stuff on E-V's Step-down mode EQ. The above mention about high-pass filter and tuning half octave below are in fact correct. B6 alignments are attributed to Thiele by both E-V and Keele, however refined by Keele while he worked at Electro-Voice.

E-V's "step-down mode" is an approximate B6 alignment, but a tried and proven concept. To my knowledge, no other major Manufacturer has used these as extensively as E-V. I would then be tempted to suggest you follow the way E-V does it since its a known recipe.

No time now, but I'll be back tomorrow to quote some juicy and easy to understand paragraphs from E-V's description/explanation of "step-down mode" operation.

BTW I'm not really in favor of building test boxes and never have made any to test woofers since its double the work and my experience at building many tells me the computer model is sufficiently precise (except vent length), if well made, to avoid them. Too many people try a quickly made box of shady construction, barely braced with questionable air tightness and then assume results will be the real thing. Not so, you must compare comparable stuff. If you make a test box it has to be identical (except finish) to the proposed final one so that you get real results of what the final product will be. Double the work for little or nothing in my view. Unless you make only one, well made, test and decide if you keep, if yes then you have only one left to do...

Richard

Attached below is my simulation in BassBox of the 1400nd

The assumption is a very modest amount of +3 dB room gain in the 25-47 hertz range declining gently above and below this region

In my own room there is generally a modest amount of room Gain.

I have shown the response curve in the 80 litre enclosure that had been empirically tuned for a flat response in the passband with room gain.

The result is very interesting for an unassisted design.

The point of the post is to illustrated to the uninitiated what can be done empirically to obtain the best result as far as a smooth and extended bass response with output where you need it.

F3 with room gain 32.5 hz
Max output 127.6 dB @ 32 hertz (122 dB @ 44 hz

The 2nd image is a overlay of the 80L and a 120L enclosure for comparison.

The 120L enclosure is tuned to 29hz
F3 with room chain is 27 hz
Max output 125 dB @29 hertz

In both cases I have shown the response without room gain for comparison purposes.
On the question of room gain and what is realistic it really is a case by case situation of the enclosure and listener location within a given room. But the curves without room gain are in free space which does not exist in a real domestic room where the enclosure is deemed to be sitting on the floor or at least near one room boundary.

Of anecdotal comment many users are so used to excessive bass from a wall boundary that a smooth flat (realistic) bass response maybe subjectively less appealing.

Others go to great lengths to equalise the bass end as flat as possible.

The spread in the tuning frequency is only 29-32 hertz and this is quite acceptable.



I would recommend as stated elsewhere the user attempt a test enclosure to subjectively evaluate any enclosure options before proceeding with the final design.

Perhaps one point to consider on higher tunings is the inevitable hump in the responses due to room gain

I will post a simulation of this later

Hi Didier and Rob,

Here is how Electro-Voice describes, in simpler terms than D. B. Keele, the "Step-Down" operation (lower tuning than normal + EQ) to get lower bass from a smaller box:

"... the "step-down" mode, which approximates a B6 Thiele alignment. Step down is a good way to extend system low-frequency response by increasing amplifier power at certain frequencies instead of enclosure size.

In step down, the enclosure is tuned to a lower-than-normal frequency. This increases system output at the new tuning frequency and reduces output slightly in the region of original tuning. The smoothly falling response which results can be equalized to provide a new system 3-db-down point that is about 0.7 that of the original. To achieve a similar response extention without equalization would require an enclosure at least twice the size.

In the step-down mode, system input is increased at the lower end of the operating range, by a maximum of 6 db close to the new tuning frequency. Note that this boost does not affect system instantaneous peak output, which is related only to the speaker's linear displacement ability and effective diaphragm area."

The frequency where an increase is applied (EQ) is called a "peak-boost frequency". According to E-V, other frequencies sufficiently close, within + or - 5%, will provide no significant performance compromise.

With regards to Subpassband speaker protection (the other side of the coin) here is what E-V says:

"Below the enclosure tuning frequency, cone excursion increases rapidly. Since acoustic output is also falling rapidly, there is no utility in driving the system with signals much below the tuning frequency. (...). ... high output systems, especially subwoofer systems, should be protected by a high-pass filter with a 3-db-down corner frequency of about 0.8 the enclosure tuning frequency... Below the corner frequency, a rolloff of 12 db per octave is usually sufficient."

So, an amplitude boost followed by a bandwidth cut at the very low end, that E-V calls "boost-and-cut equalization". "The equalization has no effect on the large-signal output but does have the effect of separating the small- and large- signal curves except at the very top of the frequency range shown."

Quoted from: Electro-Voice, Pro Sound Facts, #7, oct. 1984, pages 3 and 5

See Figure 3 in the XEQ-2 crossover/equalizer data sheet to look at examples of increases applied at various frequencies (+ 6 db with a filter Q of 2).

E-V did not mention, in this context, the LF room gain possibility based on speaker placement.

More interesting stuff to come from E-V and D. B. Keele, stay tuned!

Richard

Dear all,

I have to do some remorse here. I checked the data a little more thorough and there is real difference between LE1400H and 1400Nd, albeit having the allegedly the same cone. I have never had the 1400Nd so my findings were based solely upon the LE1400H.

Attached is a TS summary of the 14” squarcle drivers that I can find data on. LE1400H has a higher Fs and lower VAS which may explain the differences. And mine were tuned to 32Hz not 27Hz as stated Before.

Maybe someone here can fill in the blanks, that would be interesting

That means that the 1400Nd probably do need a bigger box than LE1400H to produce a nice curve. Sorry.

HoweverI do still stand by the recommendation to do real life tests before commissioning final cabinets as there are many moving parts here. Placement and shape of vents do matter and well-built test box is not much work/money compared to living room standard final cabinets.

And of course using lousy leaking test boxes is useless but as you can totally discard the finish of the test box there is no reason not to use lots of glue and insert real bracing. I typically build slightly larger boxes and experiment with volume buy introducing volume consuming object in the box or just install a shelf to make it smaller.

The EV “step-down”alternative look real interesting, I need to dig deeper into that.

-Still learning :-)

Kind regards
//Rob

Attachment 79691

Before proceeding to the next instalment of E-V/Keele "step-down mode" LF increase, I must say I disagree with Ian's following statement in post # 33:

"Perhaps one point to consider on higher tunings is the inevitable hump in the responses due to room gain"

Higher tunings don't necessarily lead to a bass bump with or without room gain. Excessively high tunings combined or not with more than one boundary (half-space or 2 Pi) speaker placement would lead to a bass bump, however this is NOT inevitable as Ian mentions.

The boxes I modeled here for Didier had quite reasonably flat response, even in the bass range, and "discrepancies" in response were noted (nothing to pull one's hair off).

As for room gain, well, one more time, if you read the help files in Win ISD Pro you will see that driver radiation into half-space (2 Pi) is assumed when modeling a driver with software using T/S parameters. In other words, one boundary speaker placement is already "built-in" the calculations/modeling:

"The data calculated from the Thiele-Small parameters refers only to the LF performance of the system; ... Further, the analysis assumes that the loudspeakers will be locating adjacent to a single reflective boundary, such as the ground, or a wall. " (P. 106). "Conventional LF alignments using Thiele-Small parameters assume 2 Pi radiation." (P. 295). John Eargle (JBL), Handbook of Sound System Design, ELAR, 1989. BTW "The author wishes to acknowledge John Hoge, who studied the outline and manuscript of the book and made many helpful suggestions." In addition to Eargle, John hoge is also a well-known Loudspeaker Engineer (CTS, JBL, etc.).

What does that mean? For normal or usual speaker placement adjacent to ONE boundary, there is no additional room gain than the one already given from full space to half-space (i.e. the usual home situation). Therefore, we're far from the "inevitable bump in the response due to room gain". The addition of a second boundary (e.g. wall) or third boundary (corner placement) could cause that along with excessively high box tuning or not...

The probabilities of excessive room gain in normal or usual speaker placement situations appear somewhat thin: "The actual home loudspeaker environment is apt to be somewhere between a 4 Pi and 2 Pi condition,...". John Eargle, Loudspeaker Handbook, Chapman & Hall, 1997, Aspects of the Home Listening Environment, P. 291. This would indicate LESS initial room gain than theory suggests (i.e. +6 db from full space to half-space), so again we would be a little further from too much room gain...

Finally, Martin Colloms (author of High Performance Loudspeakers, 1991) seems to confirm Eargle's last quote above: "... if you accept my contention that for typical rooms there is a progressive averaged bass lift of between 4.5 and 5.5 db/octave below 70 hz than a free field, ...". Speaker Builder, 4/92, P. 72, which is a little LESS than theory suggests. He wrote that in the magazine the year after his book. Regards,

Richard

Hi Richard

Thank you for your reference to my post.

I will post my simulation later in reference my room gain which is based on the Collom room gain curve.

At any rate GT did incorporate the banana curve in the statement systems for a reason.

Perhaps you can justify why that’s not the right approach?

Quote:

---

Originally Posted by RMC

Hi Didier,

I just did a modeling with a single 1400 ND in 115 L. box tuned to 28 hz in Win ISD Pro 2016 (with the usual QL 7 box loss assumption instead of the default QL 10 too optimistic number). The results are better than 1400 Pro, but still show a dropping LF response from 200 to 30 hz: about 0db@300-400hz, almost -2db@100hz, -3.5db@50hz, -5db@30hz and F3 at 60hz.

Tuning the same box at 38hz would cure the downhill bass response and put the bass level around 45hz in line with that above 200hz (except for a drop of 1.5db around 75hz) and gives you an F3 at 36-37hz or so. Not bad.

I'll be back tomorrow (no time now) with a simulation with a smaller box of 85 L. (3 cu. ft.) with 1400 ND.

Richard

---

Hi Richard,

As promised attached is the simulation in Bassbox5.1

There are perhaps a few points to cover but I will summaries as follows:

1. The 1st attachment is your tuning above with your recommendation of a 38 hertz tuning in 4.1 ft3 enclosure. I chose a thick line curve trace so its quite easy to see.

2. The 2nd attachment is the 28 hz tuning 4.1 ft3 per the bottom bass enclosure in the S9500

Ql 7
Normal fill

You can see the tuning in both simulations.
I then added the Collom's Room curve.
This curve is a defacto room gain curve that easy to find on the www. Collom's does cover the subject well in his book but its not something to loose sleep over.

As pointed out earlier this room gain stuff is an assumption but assumptions are useful when attempting to create a design that will perform well under several room placements in a domestic environment. Not on an infinite flat baffle.

Generally most users who live with the WAF have to place their JBLs near a wall boundary, not 2 metres out from the wall and 2 metres from a side wall like some audio reviewers with a financial interest who like publishing magazine curves.

The Japanese market and living room space being at a premium would appreciate "room assisted designs" and this seems to be JBl's pitch.
The statement systems are not Texas Bookshelf systems.

1. The hump I referred to is quite visible. Its not false news and its right where you chose to tune the port.

2. Putting aside the room gain I would personally not recommend a 38 hertz tuning with any JBL extended bass woofer for domestic purposes simply because it not appropriate. Such a tuning would leave the system unloading on popular music program material without a high pass filter immediately below the port turning. The user my not have the means to arrange sic a filter.


3. As far as my own research goes JBL used the lower enclosure in the S9500 for the S7500.

There is commentary on the www that JBL later improved the bass in the M9500 but no specific details are available that I am aware of.
JBL apparently designed two versions of the 1400nd and went with the one that sounded the best.

What is clear and its often the case with JBL, this particular driver was specifically designed for the S9500 (and S7500) by DB, not just for any old diy project. The S9500 was specifically tuned for the Japanese market to provide a faster bass response with minimal decay...a true audiophile offering to this day. The engineers in fact point out the bass is not typical of the 4344 series.

This therefore leads me to feel it would be wise for the user to trial the 4.1 L 28 hz tuning in a test box and see of he likes the sound rather than try and shoe horn the driver into doing something it was not designed for.

In summary I hope the original poster finds this useful.

Edit

A useful link:
http://www.allegrosound.com/JBL_K2_S...egroSound.html
Refer to Jbl spl response measurements
http://www.allegrosound.com/JBL_K2_M9500.pdf

Hi everybody and thank you for the interrest you have for my project. ;)
To be honnest, my technical knowledge are far from yours, but be sure that I try to understand a maximum :D
At first, I would rather built directly the "final" enclosure... with the good volume. It's already a lot of work and money and unfortunately I have not enough time to do everything.
About the size, I already know the width, as my horn is 44cm, and more or less the height if I want to be in the axis of the horn, something like 1m.
So, the possible difference, depending on the volume I need, is the depht, and I don't think WAF is a big problem if the difference is 10 or 15 cm ;)
Now that I know that I will use the 1400Nd (As a 12 inches is not abble to do something so good in a smaller box and that I have the speakers at home) what I only need is the best balance betwen volume and performance. (but I guess everybody undestood :D)
To give you an idea, the design is this one (except for the colour because I didn't choose betwen modern and vintage):
Attachment 79694
For the tests, I already have the enclosures of my JBL EN5, 140L, a possible solution is to to fit a part of wood on the baffle to fit the 1400Nd instead of the 2234H, filling the volume with 20L of something?
Here are my JBL EN5:
Attachment 79697

Hi Ian,

RE: "GT did incorporate the banana curve in the statement systems for a reason.Perhaps you can justify why that’s not the right approach?"

Its quite convenient and easy to say left and right that GT did this or that, must have tought such, etc. like the ventriloquist does with the puppet, knowing GT isn't here to confirm, deny or even explain what was/is on his mind. Not much more than hearsay or rumours then. I sure don't see him around here.

Moreover, the sad reality is that GT, though a quite qualified fellow, is an Engineer that wrote very few technical stuff to share his knowledge with others... Since the mid seventies I'm into Audio, have a well stocked library and I know of only two articles none by himself(!), one he wrote with John Eargle in 1986, and one he wrote with Lorr Kramer in 1983 on 2245H subwoofing, which I have.

Did JBL incorporate the banana curve in all its speaker? Very doubtful. Then, what was the specific goal, what were the constraints and technical issues to solve, what is the target market/intended purpose and use of the product? What is the intended speaker placement in the room? etc. Facts matter. References too.

Richard

This is dismaying to see an attempted offer of help to a newbie descend into a squabble between two ambitious amateurs, one of whom tries to demean a person who in these forums is an esteemed guide and honored guest. Just because you don't see GT here does not mean he is not here. If you are not an amateur, please flesh out "into Audio"; give us all the details of your portfolio--that is, all of your successful speaker designs that have been done for a real speaker company who has invested money in tooling, manufacturing and marketing the design successfully. I think I already know the answer to that, because I believe an actual accomplished engineer knows better than to trash talk the work of another accomplished, widely recognized engineer. If I am wrong, I apologize in advance. We like this "fellow" GT around here, so please, no more of this kind of dismissive remark directed against him.

Quote:

---

Originally Posted by RMC

Hi Ian,

RE: "GT did incorporate the banana curve in the statement systems for a reason.Perhaps you can justify why that’s not the right approach?"

Its quite convenient and easy to say left and right that GT did this or that, must have tought such, etc. like the ventriloquist does with the puppet, knowing GT isn't here to confirm, deny or even explain what was/is on his mind. Not much more than hearsay or rumours then. I sure don't see him around here.

Moreover, the sad reality is that GT, though a quite qualified fellow, is an Engineer that wrote very few technical stuff to share his knowledge with others... Since the mid seventies I'm into Audio, have a well stocked library and I know of only two articles none by himself(!), one he wrote with John Eargle in 1986, and one he wrote with Lorr Kramer in 1983 on 2245H subwoofing, which I have.

Did JBL incorporate the banana curve in all its speaker? Very doubtful. Then, what was the specific goal, what were the constraints and technical issues to solve, what is the target market/intended purpose and use of the product? What is the intended speaker placement in the room? etc. Facts matter. References too.

Richard

---

Hi Speakerdave,

You are absolutely wrong. The comment isn't directed at GT in any way, but rather at the ventriloquists here. And at the "Perhaps you can justify why that’s not the right approach?"

Richard

I found an extract of an old test report here with a frequency response of the S7500

I don’t read French so maybe one of our Canadian members can translate the comments

At least this gives an indication of the S7500 performance

If there are any concerns about copyright l will delete but for now it for all interested.

I am not going to buy into any concerns other than “inaccurate advice” or “misinformation” once posted creates confusion and at best down sells any reason someone might come here for help.

I think its far better to attempt a thorough investigation and present factual information of what Jbl actually did in a particular design as opposed to reinventing the wheel.

Quote:

---

Originally Posted by Didier

For the tests, I already have the enclosures of my JBL EN5, 140L, a possible solution is to to fit a part of wood on the baffle to fit the 1400Nd instead of the 2234H, filling the volume with 20L of something?
Here are my JBL EN5:
Attachment 79697

---

That’s one way to get started.

A few options:

You could put in a false bottom and fill the cavity with dried sand. Alternatively if you know the height for the horn when you are ready cut the bottom off the box and put on a new bottom.

You could then place a slab of granite under the box like Jbl did to get the height right for the horn.

Before you make the box permanently smaller in volume try the box un modified with a new baffle over the top of the existing baffle. Sand bags or bricks are an option.

It’s much easier to make a bigger box smaller.

PVC drain pipe is ideal for a temporary port 90 or 100 mm.

When you decide on the port size (cross section area) the length of the port not difficult to determine with the aid of a simulation program.

My wife would never allow a diy speaker in progress in the house these days. Years ago l actually built my Jbl 4345 clones inside my apartment between jobs. There were two problems. Dust and noise but it was worth it in the end.

Quote:

---

Originally Posted by Ian Mackenzie

I found an extract of an old test report here with a frequency response of the S7500

I don’t read French so maybe one of our Canadian members can translate the comments

At least this gives an indication of the S7500 performance

If there are any concerns about copyright l will delete but for now it for all interested.

I am not going to buy into any concerns other than “inaccurate advice” or “misinformation” once posted creates confusion and at best down sells any reason someone might come here for help.

I think its far better to attempt a thorough investigation and present factual information of what Jbl actually did in a particular design as opposed to reinventing the wheel.

---

Thanks for these pages, do you have the comple article ? This is from La (Nouvelle) Revue du Son, who stopped their activities in 2009, where Jean Hiraga was the Chief Editor. There are also some French natives from the old continent here but translating would require some time and effort. May be to concentrate this effort only on relevant technical information. Anyway Didier does not need any translation:D