CNC copy of an 2328 adapter

[Lee in Montreal]

I have started drafting a 2328 from an actual unit I have in hand. I understand there might be inconsistencies between units and that castings can vary widely. This is a work-in-progress and I am trying to understand the progression rate from throat to mouth.

51857

[1audiohack]

Here's some dimensions to add to your pool of averages.

I have two with an exit width of 3.915" and three at 3.850" and one with a vertical exit at 1.370", three at 1.390" and one at 1.400".

[Lee in Montreal]

Thanks for the info.

Looks like the average width could be 3.875" (3-7/8") and height could be 1.375" (1-3/8")

[Mr. Widget]

Your drawings don't seem to show the greater negative draft of the "balloon" portion just after the round entry. I believe this detail is significant and not at all considered on the simple throat plates used by Westlake on their later horns and by others.

I think your averages seem correct.


Widget

[Lee in Montreal]

That "balloon" irregularity is what puzzles me. I want to transfer the drawing into Solidworks and measure the area progression.

[Ruediger]

That "balloon" irregularity is what puzzles me. I want to transfer the drawing into Solidworks and measure the area progression.

Can You post a few detailed pics of the balloon?

Ruediger

[Lee in Montreal]

Can You post a few detailed pics of the balloon?

Ruediger

Here it goes. The horizontal progression is smooth and makes sense. The vertical progression shows the Ø 2" machining ending into a cavity (with a bad step). I believe that this "balloon" is only the result of the mass production at reduced costs where minimal machining must be involved. Considering the wide tolerances in many units, precision and accuracy were probably not the first target.

51864

[Ruediger]

Here it goes. The horizontal progression is smooth and makes sense. The vertical progression shows the Ø 2" machining ending into a cavity (with a bad step). I believe that this "balloon" is only the result of the mass production at reduced costs where minimal machining must be involved. Considering the wide tolerances in many units, precision and accuracy were probably not the first target.

51864

Maybe the designer had an idea similar to the following one.

Perhaps You have seen the weird folds in the Klipschorn, where the horn gets wider in the vertical direction, and narrower in the horizontal direction. It opens more rapidly in the vertical direction than it should, and as a compensation gets narrower in the horizontal direction.

The 2328 gets wider more rapidly than it should, and it gets narrower from the middle of the balloon towards the slot.

Let's check the dimensions. Upper frequency bound is 10 kHz (2440 driver), with a wavelength of 3.4 cm. The dimensions of the 2328 (magnitude-wise) are roughly 3.4 cm.

Upper frequency bound of the Klipschorn is 400 Hz, but let's calculate with 340 Hz, which is about the same (magnitude-wise). @ 340 Hz the wavelength is 1 meter, which is roughly the dimension of the Klipschorn.

So at the slot it's a "Klipsch-style" horn expansion. If a Klipschorn makes sense, then this adapter must do as well.

At the round entry to the adapter You can't do much to transform or guide wavefronts. Any expansion is wrong. The sudden change from the entrance to the balloon is a discontinuity which does no harm. But the balloon itself paves the way for the "Klipsch-style" expansion.

Weird idea? May be :)

Ruediger

[Lee in Montreal]

I am currently looking at all the possibilities for that "ballon". BTW I measured the height at 2.35" in both units of a paired set I have. I shall compared to anothr paired set currently in use. 2.35" was measured by cutting small wood sticks inserted in the ballon. From an older 2004 thread by widget, I understand he measured 2.55" on his own units. I suppose that perhaps using an average height would work.

Next step is to borrow a bore gauge and take measures from the inside at specific intervals and model a curve.

51866
51867

[1audiohack]

I would wager that the "balloon" is intentional. A transitional form to help manipulate a plane or spherical wave front into a fan shape. Also used on more than the 2397 so it is anyones guess to what if any particular horn it is optimized.

[Mr. Widget]

I shall compared to anothr paired set currently in use. 2.35" was measured by cutting small wood sticks inserted in the ballon. From an older 2004 thread by widget, I understand he measured 2.55" on his own units. I suppose that perhaps using an average height would work.Two things to consider:

A) I don't remember measuring the "balloon".

B) I won't vouch for the accuracy of any such measurement.

OK, and a third thing to ponder... if I recall correctly (use an alphabet salad for the phrase if you like) doesn't the Altec 511 horn have a similar "balloon"? I know some horn I was once checking out did, but I can't remember which one... then again, I don't remember posting a measurement in 2004 either. :D


Widget

[Lee in Montreal]

I don't know how to thank you, Widget. Sacrifying a 2328 to chop it up was brave ;-) :D
From the the cuts in the picture, I was able to faithfully retrace the horizontal and vertical paths. Thanks again.

51876

[Allanvh5150]

The balloon is probably just to transition from the circular driver entry point into the square exit point.

[Wardsweb]

[QUOTE=Allanvh5150;316237]The balloon is probably just to transition from the circular driver entry point into the square exit point.[/QUOTE
If anything like fluid dynamics, so the sound bubble exits on both the vertical and horizontal at the same time.

You see the bowl cut a lot in head porting for race applications. It is just the reverse, square or rectangle to a round valve.

[Lee in Montreal]

I forgot to mention that my intent, unlike what the title suggests (my bad), is not to produce copy of a 2328. I am doing the exercise of redrawing a 2328 in order to better understand its principle. My goal is to design a 1" thick aluminum adapter as I seem to understand that the shorter adapter will be a bit better at higher frequencies. That adapter will work on 2350 and 2397 horns. I will make a few units for personal consumption for 2" throats, but the design can be adapted for 1" and 1,5" throats.

[Lee in Montreal]

I will have a 3D model in a few days, as well as a chart displaying the progression of the cross section from throat to mouth. Adapter is 1" thick for 2" drivers. It's more in line with a Westlake style adapter than a JBL one. No "ballon" here... ;-)

51905

[Wardsweb]

I was working on something very similar today. You can see the adapter block on the left of the horn. The black is spray paint I used to match the horn throat. This gets cutout to transistion to the 1" cutout. It is 1" thick. For comparison to it's left is a 3/4" adpater I got from Bill Martinelli years ago.

51906

[Lee in Montreal]

Nice.

Are you going with a 1" driver? Also, after prepping, assembling and finishing the horn, I found very easy to fit the fins from the throat. Carefully slide them in. There's just enough play for an easy slide. And as the fins are razor sharp, you won't scratch the horn by fitting the fins from the thoat.

Can't wait to see your finished horn with the mirror-like polished fins. On the last stage of aluminum polishing, I like to use a rubbing compound called Autosol that leaves a mirror finish.

[Wardsweb]

Yes, a 1" driver as that is what came in the C50 speakers I bought. I may go 2" later.

I use white tripoli to polish and Happisch Simichrome to mirror.

[reVintage]

Of course the baloon is intentional as said earlier. The flare is 160Hz exponential. This is what the WE594 had. JBL copied it when they made their 375. For some reason they changed the flare to 180Hz.
Also the 2328 differ significantly from the prototype according to the designer. Think we had that up to discussion in an earlier thread about the adapter a few years ago.

http://www.audioheritage.org/vbulletin/showthread.php?15654-JBL2328-adaptor&highlight=2328

[Lee in Montreal]

Hi Lars

The ballon has now been integrated into the draft. Interestingly, I checked another set of 2328 I had and the machining at the thoat is very different. One set is smooth, one set is notchy. The notchy set is what made me think the ballon was an "error". While it is obvious wuth the smooth set it was designed in. I guess I now have a complete draft that I can use in a few years when most 2328 will have been melted or junked :D. That project has evolved into a 1" adapter.

[reVintage]

The thing you draw in post 16 is totally wacko. If you want to add a 1" to 2" adapter let it be 160Hz and something like 200mm long. If you make it 400Hz(like most 1" drivers flare) it should be something like 100mm.

Best is to check the flare and length from a radial horn spreadsheet. 2397 is a cut radial. Yggdrasil have one and I also have one.

Also the prototype probably was 4"*1,5". Have checked my 2328´s and it seems OK.

The fins should not be polished as we have low airspeed. Glass blasting should be fine.

[Lee in Montreal]

Hi Lars

You are bringing a few points that deserve to be discussed, even though I don't entirely follow you. ;-)

1" - 25.4mm, therefore if I design a 1" adapter, it won't be 200mm. The goal is to make a compact adapter while everyone understood that the cut-off will be higher.Therefore the lower cut-oof will probably be around 1khz and not 800Hz. But it is fine with me as it is what I am after.

As far as boundary layers caused by a polished/smooth surface, this is a well know/documented subject in automotive and aeromotive industry. Air and fluids do stick more to a perfectly smooth surface than on one with irregularities. It slows down the flow by creating unwanted friction because air and fluid will bond to the surface. But in reality, is there any airflow in a tweeter or a medium driver? Does the membrane move enough to create flow, as such is the case with a woofer? And even with a woofer, if I use a smooth ABS pipe, will it have detrimental effects?

[reVintage]

Hi Lee,
You are probably right in that the air speed is so low that the flow is almost absent. Still in low speeds like in the Tornado catamaran I used to sail, rudders worked best when matted(?) down. But there speed was higher, up to 30 knots:D.

About 20cm it is correct when you do the 1-2" adapter with the same flare as 2328. very simple! As I also said the 400Hz expansion would give you ca 10cm but the they don´t fit so well together. Better to do a 400Hz "2328" in that case. Hornresp. is a very good tool.

About the bulb you get the answer when you check Altec 511 but the "bulb isn´t smothed out there. Actually it is what happens when you follow the exponential expansion rearwards you come to the point where you come to the square point. If followed rearwards the horn form would be vertical. So cut the radial at the square point to go from square to round.

[reVintage]

Found the pic´s of a conversion I did for fun to the 511, 2328 and 2" driver. Still haven´t tried them though.

It is not completely correct as the 2441 is 180Hz, the 2328 160Hz and the 511 270Hz.I actually cut the 511 where it had exactly the same opening as the 2328.

[Ruediger]

You can join segments of any flare (exponential, hyperbolical, conical) and of any cutoff frequency. The segment with the worst properties determines the behaviour of the combination. If that segment is short compared to the wavelengths involved, then it's a simple discontinuity.

If You wanted to achieve plane waves so that You could more easily design an acoustical lens, then You would pick a very low cutoff frequency.

Ruediger

[reVintage]

Hey Ruediger,

Off course you can mix but to get an optimal result an uniform flare all the way is the way to go. About 2397 style horns I think it is a big mistake to addd the driver directly to the hornthroat. There must be a adpter with a"bulb" to change the flow from round to radial.

So if wanting to go 1" and a crossover of 800Hz, a flare in the ballpark of 400Hz all the way should be great.

But it can pose problems when a driver like 375, 2441 etc. has the built in horn with a flare of exponential 183Hz.

I designed and custom-ordered Le`Cleach horns 200Hz/T=0,8 from Auta-Tech as that was very close to the initial flare of the 2441.

[Ruediger]

... to get an optimal result an uniform flare all the way is the way to go ...

... There must be a adpter with a"bulb" to change the flow from round to radial ...

... So if wanting to go 1" and a crossover of 800Hz, a flare in the ballpark of 400Hz all the way should be great ...

... But it can pose problems when a driver like 375, 2441 etc. has the built in horn with a flare of exponential 183Hz ...



Hi reVintage,

1st statement: disagree.

2nd statement: that's what the designer had in mind. Room for educated guesses.

3rd statement: flare frequency = 1/2 * lowest frequency to be transmitted holds for exponential horns in general, not just for adapters. But the adapter cannot cure what the horn spoils.

4th statement: disagree.

Ruediger

[reVintage]

Hey Ruediger,


Hi reVintage,

1st statement: disagree.

2nd statement: that's what the designer had in mind. Room for educated guesses.

3rd statement: flare frequency = 1/2 * lowest frequency to be transmitted holds for exponential horns in general, not just for adapters. But the adapter cannot cure what the horn spoils.

4th statement: disagree.

Ruediger 1st So a horn shouldn´t have the same flare from throat to mouth? Totally disagree or is this a missunderstanding?

2nd Would have been interesting to see the original prototype mentioned in the other thread. Did measure(long ago) at two points except mouth and throat and and they indicated 160Hz.

3rd. Yes, 1/2 lowest in general is correct and I even speak of the horn inclusive driver by choosing this frequency. Both Altec and JBL has ca 400Hz flare inside. The adapter is a part of the horn and should be counted in as part of it. So it should of course have the same flare(st1 above). This is what is ment with "all the way".

4th. JBL published figures about 160Hz flare. Probably they got that from the sketch made by JBL around 1950. It was the a direct copy of the WE594 flare. For some reason the internal flare where made ca 2cm shorter(with same At and Am) in the 375. So if you take apart the driver it is easy to measure and do some math to find out it is 180Hz with exponential flare. This is a fact and can easily pe proven. It is no problem to measure from outside too. Ballpark figures are L=69mm, Am=20cm2 and At=11cm2.

About the 2328 one must remember that the first part of the radialhorn(2397) is in the adapter. So doing a 1" adapter based on guesses is not the best idea from my point of view.

Using Yggdrasils spreadsheet for a 160 exp radial the figures aren´t to far away for the adapter. From Lees nice drawing it isn´t clear if the bulb is at five or six cm´s from the throat. Haven´t mine available as I am at work.

http://www.revintage.se/160.xls

EDIT: Some second thoughts. If making new 1" adapters for 2397, why not go for 2397 flare?

[Ruediger]

Hi reVintage,

I did not say the horn should have a varying flare, but rather it may have a varying flare.

See the Altec Manta Rays, and the JBL true Biradials. It gives You some freedom in designing.

I wrote "educated guess", not "guessing".

Ruediger

[reVintage]

Hey Ruediger,

OK, but "may" could go for any horn, good or bad. Still in my opinion the best horns for HiFi today are with the same flare all the way. About the biradials I don´t know as I haven´t examined them, but they sound OK. Would be nice if you showed their design.

Did some Googling and found they are CD horns. Keele did the EV HR series and after that, at JBL in order to get around some patents, designed the Biradials. Looks nicer to me as they are smoother without the edge(?) when going from exp to conical. Read that paper long ago so I don´t remember the explanation.

Can´t find the word "guessing" in my answer to you;). I mention "guesses" after my answer to you, about a subject that wasn´t directed to you mainly.

I used it after having seen 1" adapters here with designs that must have been based on guesses.

[Lee in Montreal]

I used it after having seen 1" adapters here with designs that must have been based on guesses.

Did people with their 1" thick adapters pretend it was doing the same task as the 4" long adapters? ;) If they did, well, that was most likely a wrong pretention. If they acknowledged a higher cut-off, then what's the problem? A different tool for a different purpose.

[reVintage]

This is how JBL adapts 1" drivers to 2397. As the adapter is short we can assume it has a higher frequency flare.

52018

And this Mr. Widgets idea, probably you will loose bottom end but gain top end, question is what happens to horizontal dispersion:

[Lee in Montreal]

Hi Lars

I am confused by the use of the term "1-inch adapter", and perhaps this is why I don't understand how this topic is evolving. This thread was initially about a 1" thick adapter, not an adapter for a 1-inch driver. I guess I'll try to do my best to avoid any confusion in the future by using more precise terminology.

In regard of the piggybacking of adapters for a 1" driver (as pictured above), doesn't it bring detrimental effects in the upper frequencies? :p

Anyway, let's keep up the discussion as we'll all learn something...

[reVintage]

OK, lets start again. Did you mean a 1" thick adapter for a 2" driver on 2397?

Flare again;). If we consider your adapter as being with an exponential progression it has a flare of just below 600Hz.

[Eaulive]

:lurk::D

[Lee in Montreal]

OK, lets start again. Did you mean a 1" thick adapter for a 2" driver on 2397?

1" thick adapter for a 2" driver to be used on any 3.5" x 3.75" mounting pattern. Be it a 2397, 2350 or any Westlake or other other.

Lee

[reVintage]

As you see above it has a flare of about 600Hz.

[ivica]

The balloon is probably just to transition from the circular driver entry point into the square exit point.

If linear horn surface expansion ("parabolic horn") is the goal, remaining horizontal conical flare, then such shape ( in vertical curve) can be expected.

look at the:

http://www.audioheritage.org/vbulletin/showthread.php?15654-JBL2328-adaptor&p=213095&viewfull=1#post213095


Reagrds
Ivica

[bldozier]

:shocking:

did you make these?

[ivica]

Hi bldozier,

Yes, You can see on some of my old post the whole horn made as a kind od DD66k inspiration.
As I need not so wide horn i have recalculated horn lips to be about 60cm wide and about 25cm height.
I have made two type of horn mouth, one for 2-inch driver (I have 2446, and 2450), and 1.5-inch driver (I have 2450-1.5-SL driver).
With 1.50inch throat version I have get better horizontal dispersion ( about +/- 45 deg), while with 2-inch throat such hoizontal
dispersion is about +/- 35 degs.

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=60299&stc=1&d=1381156034
http://www.audioheritage.org/vbulletin/showthread.php?34180-2445J-with-2311-Horn&p=353522&viewfull=1#post353522

regards
ivica