2426 2445 tapered throats

[tagerbaek]

Hello JBL savants,

I'm putting my 2426 and 2445 compression drivers on home-made wooden tractrix horns.
The throats of the drivers are tapered: on the 2445 it's visible, on the 2426 the mesh hides it.
I need to know these angles in order to calculate the tractrix horns. Can anyone pls help me?

TIA Thomas

[Robh3606]

Well if the 1-2 inch adapters just follow the same flair rate?? Why not grab a 2327 adaptor and measure it that way.


Rob

[tagerbaek]

Thanks for the input, ROBH3606 - Alas, if only I had a 2327.

But: the 2327 is 105 mm long. According to Pythagoras, if it were straight-sided, then its angle would be 6.56 degs, to convert between 1 and 2 inches.
However, I have measured the 2445 exit to 5 degs.

And: the 2445/2426 drivers didn't exist when the 2327 was designed, but the 2440 did, which had a straight throat, AFAIK, maybe I'm wrong.

So: the exact exit angles and curvatures of the 2445/2426 are still needed.

Because: it's a bitch calculating and making a tractrix. Trial and error is not an option. Best get it right from the start.
TIA Thomas

[frank23]

as fas as I know, the 2445 has the same phaseplug as the 2440 and can use the same diaphragms and has the same throat 2" exit

in fact I looked it up and the 2445 brochure states "phasing plugs and throat exit tubes are identical to previous alnico models"

a suggestion for measuring might be the following method:
- choose a [plastic?] tube with an outer diameter larger than the end-inner diameter of the throat exit, but smaller than 2"
- insert the tube into the driver throat
- measure how deep it goes in when gently pushed in
- calculate the angle from the depth were the diameter is x as referenced to the 2" exit diameter

how come this reminds me of a certain movie?

frank

[Robh3606]

Hello Thomas

Been thinking about this. Why do you need the flare rate before the 1 opening?? You can take any 1 driver and drop damn near every kind of horn known to man on it from different manufacturers of both horns and drivers to boot. All those horns could not possibly match the rate inside the driver. They all have their own depending on the type. Look at a exponential like a 2307 vs. a 2370 biradial. Night and day. So what's the poop???

Rob

http://www.lansingheritage.org/image...l300/page2.jpg

[tagerbaek]

Hello guys,

frank: OK, I was wrong wrt the 2440, it's tapered allright. I have already measured the 2445 using a steel ruler and a graded drawing arc, this is reasonably accurate. However, this I can't do with the 2426.

rob: you're right about the horns, but that's just the point.
I have a pair of 2370s and a pair of large aluminium exponential horns; they sound different (you can hear this just by putting them to your mouth) and are definitely not made for home stereo usage, they are optimized for high SPL PA.

That's why I've decided on tractrix horns, which are much more suitable for home use: they present a near optimum load to the driver, they are free of resonances because of their rapid expansion, and they present a perfect spherical source with no edge artefacts, ie, they are smooth and not horny.
This at the price of a relatively large, circular mouth, lack of gain, CD, etc.

But, if there is an abrupt change in the flare rate, the acoustic impedance will at that point also exhibit a discontinuity, leading to resonances, reflections etc.
That's why the driver exit angles are important.

I found a picture of a cut-through LE85/2420 on a 23xx exp horn, showing the driver and horn forming a perfect continuum. If no better info turns up I will measure the angle from this photo.


BTW it's gonna be 136a, 2118j, 2445j, 2426h, 2405h in a 140 liter net (as seen by the 136a) cabinet, tuned to an fc-3dB of about 32Hz; 5-way active, pure class-A; fpga-based digital xover.


PS I still need the 2426 exit angle.

TIA Thomas

[Robh3606]

"But, if there is an abrupt change in the flare rate, the acoustic impedance will at that point also exhibit a discontinuity, leading to resonances, reflections etc.
That's why the driver exit angles are important."

I understand why your are asking just not sure of the need. I have heard tractrixes and they do indeed sound good. That said you can purchase tractrix horns from Dr. Bruce Edgar of Titan fame. I have heard his horns, both the 1" and 2" versions. When you purchase the horns as far as I know there are not multiple versions to match up with internal driver flare rates. You drop on your driver of choice and go from there. I have never heard of someone doing that just thought as long as the opening matched you would be fine. I am curious about this as the internal flair rate could be conical as an example. If this is the case if you continued the flare rate wouldn't you end up with a conical?? How would you transition between the two without running into the problem you are trying to avoid??

Rob

[Zilch]

The little screens come off in their circular frame.

You may have to purchase a replacement from JBL Pro parts, is all.

Hard part will be finding the part number....

[tagerbaek]

Hello guys,


thanks for the input,


rob, I'm arriving at the same conclusion as you, regarding the horn/driver angles: most people simply blissfully ignore the problem, and sometimes it sounds good, sometimes not, but they're none the wiser.

This wouldn't be of any great concern for me, were it not for the fact that horns are non-tweakable. It's a one-shot deal, in particular because I'm integrating the horn and baffle.

Because every driver has a given exit angle and exit diameter, the attached horn size and flare rate are predetermined, even if you are only matching the size and angles (and not the higher-order derivatives of the curvatures); this goes for exp, con, tractrix and all the other simple expansions.

For example, matching a tractrix horn to a 2426 style driver yields a mouth dia of 1inch/sin(5 deg) =11.47in;
and a length of 11.47inch/2 x (ln(cot(5 deg/2))-cos(5deg))=12.24in.

A tractrix on a 2445 driver has twice the diameter and length.

The wife acceptance factor dictates that these horns are too large. I want half the size for both horns. This means starting off, at the driver end, with a flare rate which is half that of the mouth flare rate, increasing the flare rate towards the mouth, while actually exceeding the mouth flare rate somewhere along the way to make up for the initial slow expansion.
So, obviously I'll have to depart from the perfect tractrix, but what would be the best compromise here? Putting an exp or con section between the driver and tractrix horn has been suggested elsewhere, but in the present case would only exacerbate the problem, as the tractrix is the fastest expanding contour.

This is a construction project, not an academic exercise.
I need, and would very much appreciate any input which could help me make the right decision regarding those tractrixes.

Sounds complicated? Yes.
Am I overcomplicating things? Maybe.
Am I the first person to have considered this issue? No way.




Cheers, Thomas


PS, at my old university the joke went:

the M.Sc. builds a machine; it doesn't work, and he doesn't know why;
the Ph.D. builds a machine; it doesn't work either, but he knows why;
the B.Sc. builds a machine; it works, but he doesn't know why.

[Robh3606]

Hello Thomas

I certainly can't help you with this however Dr. Edgar certainly could. Check out his site and try contacting him. I doubt there are many others who are more knowledgeable and he is. He is friendly and a very easy person to converse with. Try a PM to Steve Schell as well. He is a member here and may be able to suggest the best course of action. Dr. Edgar has tractric horns designed for the 2441 and they sound great! Take a look at the Titan or Google search for the Titan as the site is not finished. His 300hz horns are not all that large. He sells them unfinished and they could be integrated into a baffle easilly as you are planning as opposed to turning your own. The attached links give you photo's and his email address. You might want to try a post in the High Efficiency forum at AA. There are several guys there that have scratch built and maybe they could help as well. Have fun and please post some pictures when you get things going.

http://www.theshowlasvegas.com/main/...dec80d33e05b3d

http://www.dagogo.com/CES2005/CES2005-C-E.html

http://www.robbreportcollection.com/...dern-Sound.asp
Rob

[subwoof]

tagerbaek:

I have a 2420 with a very weak magnet that I use to teach sound engineers how a driver works. The throat is in excellent condition and if you PROMISE to return it I can send it and a 2440 throat to you for measuring with the understanding you will post the results for the heritage members to use.

sub

[Earl K]

Hi Thomas,

(A) Small Format :
- I'm under the impression that the 2425/6, along with virtually any 1" exit driver from that generation & earlier , used a 350 hz , exponential taper starting at the phase plugs' termination & finishing at the 1" exit .
- That taper just happens to look conical since it only represents the first couple of inches from a pretty large horn.

(i) Unknown , is the "flaring constant" or "m factor" for this 350hz horn . It is most likely in the 0.20 area . Also unknown to me is the actual starting diameter of this throat taper / ie ; the diameter of the horn, right at the phase-plug end of the taper .

(ii) This flare constant ( actually, 0.35 on down to 0.10 ,with 0.20 "looking" the most correct ) give initial throat tapers that are very close to overlaying ontop of each other / and most importantly / "appear" as straight conicals for the initial, incremental expansion of the horn flare. "Appearing as a conical" being the operative word. I don't know what Altec, JBL or others where actually able to manufacture given the technology of the time and their perceived " return on investment" . FWIW; I believe they simply used a conical taper since it is also very close to the lower numbered exponentials ( flare constants ) which I've just mentioned .

- Newer, more modern drivers must be using a different "flare constant" to achieve the rapid expansion dictated by their very short/shallow throat sections, travelling through these "thin" drivers . The "m" factors "look" to be in the area of 0.50 to 0.60 .

- BTW ; an exponential flare constant of around 0.25 to 0.30 ( for a round mouthed exponential horn ) gives a horn that would be hard to tell apart from the typical "tractrix" sorts that I've seen advertised .

(B) Large Format :

The initial taper inside the driver ( Altec 288, JBL 2441 and TAD 4001 ) are all based on a 225hz exponential horn . Everything I've said above, also applies here. ( ie ; "Flare Constants" aren't known [ at least to me ] / & the observations of older, deeper designs versus newer shallower drivers ) .


- This info is derived from two sources :

(1) The New Audio Cyclopedia / also called the "Hand Book For Sound Engineers" , published by Howard H Sams ( Sams' reference #21983 ) . page 469 . The info on the initial throat tapers of older "classical drivers" is found there. Throat taper is called fpg and is referenced within a comparison chart about some relative 2HD values ( second harmonic distortion for the mentioned drivers ) .

(2) The Sound Recording Handbook, published by Howard H Sams ( Sams reference #22583 ) read pages 179 to 191 for info on horns and most importantly their flare constants ( or "m" factor ) .

Cheers

ps : by all means, contact Dr. Edgar and ask him his thoughts on all this.

- Okay, I now see subwoofs pics. He can measure the initial diameters of the tapers found back at the phase-plug .

- In case it's not obvious by my post. This discussion can also be framed as a question of ;
" How important is the matchup of two different Flare Constants within the first 1 to 2 inches of any horns initial expansion ?".
- You might ask the good Doctor for his perspective on this .

[Maron Horonzakz]

Rather than futzing with the conacle angle of the throat on the 2446...Go to the 2447 or 2451. You eliminate the small conacle cone and go directly to to the phase plug and couple directly to the horn flair of your choice. The TAD 4003 also does this & Tad 4002 front conacle detaches to bring phaze plug closer to horn throat. All then start with a 1 !/2" opening & calculate horn Flair to infiniti. You also will be starting with the newer driver designs.. This all brings you a couple inches closer to couple driver diaphram/phaze plug. to horn tracktrix eliminating the distortions you claim are there. JBL has a engineering white paper on this subject & how it reduces some distortion problems. That what the new 1 1/2" throats are all about. Go for it...Stop futzing mateing old driver designs to new technolagy.

[tagerbaek]

Hello there,


rob, thanks for the links and your interest, I'll also try madison and rec.audio.tech before bothering Mr Edgar.
About the horn size: the cabinets are already built, I can only just squeeze in those 5 drivers. Got carried away, I guess! Also, I'm already looking at a combined weight of the whole thing of +100kg, so larger horns are out of the question, I'm afraid.


subwoof, that's very kind of you, but I live in Munich , Germany (born in Sweden, with Danish citizenship, as it were), so shipping costs are prohibitive. However, if you could get the time to just measure the inside angle of the 2426 throat exit (or the diameters of the tube at both ends) this would be greatly appreciated.


Maron, I just got the 2445's, I'm sticking with them. Voigt's tractrix patent is from 1926, I think - hardly new technology, but OK, you talked me out of it: I'll never again futz a conacle.


earl, interesting info; now we're talking. I'll check up on that. The 2445 throat does line up perfectly with a steel ruler, though, suggesting a very nearly conical shape.


I'll keep you posted as the project develops.


Cheers ,Thomas

[subwoof]

OK guy, break out the calculator....

Ich bin ein bissin "engineer"

The throat sides ARE straight inside ( conical ). The overall length of the throat is 2.025 inches ( oops....)

The throat is "press" machined and the thichness varies along it's length. The best I could do was use a thin 6" rule inside the throat and make a right angle triangle with the square. Slight thickness of the screen, glue and my aging eyeballs make for an approximation but it's close enough...

zwei bieren bitte...

sub