Upgrade: SAM1HF or 2452H-SL in ScreenArray horn?

[Jonas_h]

I am currently using a ScreenArray horn with a 2432H compression driver. I really like the sound, but I feel I am missing some resolution and extensions in the high frequencies. I have thought about upgrading to either the SAM1HF horn/drivers or getting a 2452H-SL driver for my current horn. I have heard the SAM1HF and really like it, but I have read many good things about the 2452H-SL and like the fact that it is a 4" driver which covers the whole spectrum.

All inputs are appreciated :)

[Mr. Widget]

The SAM horn with a 4" Truextent Be diaphragm in an appropriate JBL driver would likely be the current ultimate upgrade. That said, I have been quite happy withe the stock 435ALs.


Widget

[4313B]

I am currently using a ScreenArray horn with a 2432H compression driver. I really like the sound, but I feel I am missing some resolution and extensions in the high frequencies. I have thought about upgrading to either the SAM1HF horn/drivers or getting a 2452H-SL driver for my current horn. I have heard the SAM1HF and really like it, but I have read many good things about the 2452H-SL and like the fact that it is a 4" driver which covers the whole spectrum.All inputs are appreciated :)I'll second what Mr. Widget posted. The 435Al and 2431H are quite good. And you can always swap in the 2435 diaphragm if you want a Be solution (not aquaplased like the 435Be diaphragm, which is virtually unobtainable). You'd have to hassle with the ferrofluid but Doug Button gave us the exact type and amount for the 435/2435 core and RobH3606 knows where to order it from.

2452H-SL? Don't you mean the 2450SL with Truextent Be diaphragm?

[Jonas_h]

What are the difference between the 2431H and 2432H besides AL vs. TI?

I do actually mean the 2452H-SL - aquaplas damped titanium driver.

[4313B]

I do actually mean the 2452H-SL - aquaplas damped titanium driver.I'm pretty sure you mean the 2450SL with Truextent BE diaphragm but I could be wrong. :)

[pos]

What are the difference between the 2431H and 2432H besides AL vs. TI?

I do actually mean the 2452H-SL - aquaplas damped titanium driver.

The 2432H has an additonal "snout" that helps avoiding problems (sharp dips above 10khz, that get worse off axis) with short/rapid flare horns (like the PT waveguides or 2332).
But if the SAM horn is anything like the h9800 (and I think it is very similar) the 2431 would not have any problem with it.
The best choise with that horn would probably be the 435Be or 2435.
I am using a 2450SL+Be diaphragm and the difference with the 2435 is not that big on the H9800...
(horizontal directivity is better though, but with the vertical horn the directivity is quite similar, with the slot doing all the hard work)

[Jonas_h]

I'm pretty sure you mean the 2450SL with Truextent BE diaphragm but I could be wrong. :)
http://www.jblpro.com/catalog/general/Product.aspx?PId=325&MId=1


The 2432H has an additonal "snout" that helps avoiding problems (sharp dips above 10khz, that get worse off axis) with short/rapid flare horns (like the PT waveguides or 2332).
But if the SAM horn is anything like the h9800 (and I think it is very similar) the 2431 would not have any problem with it.
The best choise with that horn would probably be the 435Be or 2435.
I am using a 2450SL+Be diaphragm and the difference with the 2435 is not that big on the H9800...
(horizontal directivity is better though, but with the vertical horn the directivity is quite similar, with the slot doing all the hard work)
I'm only interested in "stock" components. So either the SAM1HF or a new driver for my current horn.

[4313B]

I am using a 2450SL+Be diaphragmIf I'm not mistaken, that should be about as close as one can get to a 476Mg or 476Be using Pro parts.
and the difference with the 2435 is not that big on the H9800Hmm.

[cooky1257]

Hmm, I use the 2452+Truextent Be on H9800(correct throat) clone at the mo' and it is a noticeable improvement over the 2435, ditto on the SAM1.
As I recall Zilch was full of praise for the stock 2452.

[4313B]

Hmm, I use the 2452+Truextent Be on H9800(correct throat) clone at the mo' and it is a noticeable improvement over the 2435, ditto on the SAM1.It should be with the Be diaphragm. G.T. expressed his reservations with respect to replacing a 435Be or 2435H with it in stock form though. I believe he felt the 3-inch Be diaphragm trumped the larger 4-inch Ti and Al diaphragms. I do remember him being quite impressed when he swapped out his 435Be's for 476Be's on the Array horns. Such swaps require network rework due to impedance, phase, and frequency response differences.

I personally am thoroughly impressed with the 435Be and 2435 on the Array horn.


As I recall Zilch was full of praise for the stock 2452.Back when Zilch was first starting out on the econowave thing we had a chat on the phone with respect to these various compression drivers. I passed along to him the notes from my several in depth conversations with JBL. At that time I expressed my irritation with the fact that the 2452 was nearly the same price as the 2450 given that it allegedly enjoyed a considerable savings in manufacturing. I really don't know what Zilch did with all that information or what he finally determined to be the most desirable combination.

If I'm not mistaken (and I very well could be), if someone is buying a new compression driver from JBL Pro in hopes of somewhat emulating a 476Be or 476Mg, I believe the 2450SL is the model to get along with a Truextent Be diapraghm. I'm not sure what the total cost of that package is. I've never looked into it. I haven't been paying a whole lot of attention to Pro for quite some time now. According to several sources at JBL, that should get one to about 80% of a 476Be or 476Mg. Who knows, maybe a 2452 is "eighty percent" too depending on the horn? I just don't know at this point. That was a long time ago.

[cooky1257]

Doesn't the 2450 have the coherent phase plug a la 476? The 2452 looks like an enlarged version of the 2435. For the money, even 80% is worth having.I thought the extra cost of the 2452 was down to the neo mag?
Frank

[pos]

The 2452 also has the 4-slit coherent wave phasing plug, albeit a bit shorter.
In my tests the 2450SL core outperformed the 2452 one by quite a margin.
I don't know how to express that, but the dynamic was more realistic on the 2450. The 2452 sounded more "fake" in this regard.
I think is has to do with the "optimized" magnet sturcture of the 2452 (field distortion?...)
I found the 2452 and 2435 both shared this trait...

Strangely the difference was more obvious on short waveguides than on the H9800 for example.

[cooky1257]

Pos,
When you say 'tests' you mean listening don't you?
As I recall you didn't actually measure the 2452, please correct me if I'm wrong:)
I've always understood the coherent wave phase plug slots to be curved cross section-the 2452/2435 look straight.

Cooky

[pos]

I did a lot of measurements too (more measurement than listening tests in fact).
I tested (measurement and listening tests) Ti SL and Ti ribbed diaphragms in both cores, but I didn't installed the Be diaphragms in the 2452.
With the same diaphragms installed, the 2452 measured almost exactly like the 2450SL, with only a slight boost in the lower range due to the smaller back cap.
The directivity behavior was also identical, and much better than the 2435 (but the slot in the array horn makes it a moot point).

I don't know what measurement could illustrate the differences I heard in the listening tests, but I imagine it is always the case with magnet-related issues (alnico vs ferrite vs field coild vs...)

[cooky1257]

I did a lot of measurements too (more measurement than listening tests in fact).
I tested (measurement and listening tests) Ti SL and Ti ribbed diaphragms in both cores, but I didn't installed the Be diaphragms in the 2452.
With the same diaphragms installed, the 2452 measured almost exactly like the 2450SL, with only a slight boost in the lower range due to the smaller back cap.
The directivity behavior was also identical, and much better than the 2435 (but the slot in the array horn makes it a moot point).

I don't know what measurement could illustrate the differences I heard in the listening tests, but I imagine it is always the case with magnet-related issues (alnico vs ferrite vs field coild vs...)

Thanks.

[ivica]

I did a lot of measurements too (more measurement than listening tests in fact).
I tested (measurement and listening tests) Ti SL and Ti ribbed diaphragms in both cores, but I didn't installed the Be diaphragms in the 2452.
With the same diaphragms installed, the 2452 measured almost exactly like the 2450SL, with only a slight boost in the lower range due to the smaller back cap.
The directivity behavior was also identical, and much better than the 2435 (but the slot in the array horn makes it a moot point).

I don't know what measurement could illustrate the differences I heard in the listening tests, but I imagine it is always the case with magnet-related issues (alnico vs ferrite vs field coild vs...)

Hi POS

Is it possible to show some measurements with Ti SL and Ti-ribbed with H9800, ALL WITH 2450 DRIVER

Thanks
Ivica

[pos]

Hi Ivica

What kind of measurements are you interested in?
I usually don't keep my measurement: I use them and discard them (measurement are most of the time only meaningful for the one how did them, and only as long as he remember the parameters and constraints of the measurement session...).
I do however have kept some directivity measurement of the H9800 with eq and smoothing, with the 2452H (ti ribbed). The 2450SL or Be would be almost identical with EQ and smoothing applied...
I also have the 2435 for comparison.

2452H horizontal
56549

2435 horizontal
56550

2452 vertical
56551

2435 vertical
56552

[ivica]

Hi Ivica What kind of measurements are you interested in? I usually don't keep my measurement: I use them and discard them (measurement are most of the time only meaningful for the one how did them, and only as long as he remember the parameters and constraints of the measurement session...). I do however have kept some directivity measurement of the H9800 with eq and smoothing, with the 2452H (ti ribbed). The 2450SL or Be would be almost identical with EQ and smoothing applied... I also have the 2435 for comparison. 2452H horizontal 56549 Hi POS I agree with You that after EQ and Smoothing every driver response can be made ideal, but my attention is mainly focused on non-EQ and as less possible smoothing (say 1/24 or 1/12 octave). I have a project of constructing a horn 2-inch throat and shape something like H9800, on which I have plan to put 2441J -2446J - 2450J-2 with TI-ribbed diaphragm, so for me it would be of interest to see the behavior of the newer driver behaviors in the "near-the-same" conditions, just as a reference. So here is my measurements of 2441 with D16R2445 diaphragm with the horn 2311+2308, smoothed and 'non-smoothed' measurements 1-m from the horn mouth on-axis

[pos]

Given the same throat and horn, and of course the same diaphragm, all the coherent wave 4" CD should measure similar (2446, 2447, 2450, 2450SL/2451...)
One thing that seems to appear though is that the ferrite drivers do not have the dip in the ~3khz range that the neo drivers have. Maybe this is due to the bigger back cap?

[ivica]

Given the same throat and horn, and of course the same diaphragm, all the coherent wave 4" CD should measure similar (2446, 2447, 2450, 2450SL/2451...)
One thing that seems to appear though is that the ferrite drivers do not have the dip in the ~3khz range that the neo drivers have. Maybe this is due to the bigger back cap?

Hi POS

Well, that is the reason why I have asked some more detailed measurements data with H9800 horn, especially You have done measurements with 2450-1.5. That is, as I have expected 'derivative' version of 2450-2 (with short 3-inch horn embedded). I believe that this embedded horn, if driver coupled with 'fast flare expanded horn', as H9800 would behave different, especially over 10kHz.
As expected the horn 'loading' would produce different results compared with wave tube measurements,not to mention total acoustical 'picture'.

About the mentioned dip in the ~3kHz region, I have quite different results with the new and older , I can guess, heavier used diaphragms.
The older has much noticeable dip at about ~2kHz region, while the new one almost flat, all measured with 2311+2308.
May be that is some kind of aging of Ti-ribbed diaphragms, or some kind of Ti fatigue under heavier stress long time operation.

Regards
Ivica

[pos]

The 2450SL is identical to the 2451 (except for the bolt pattern): it has no integrated throat and has the phasing plug directly on its 1.5" output.

[ivica]

The 2450SL is identical to the 2451 (except for the bolt pattern): it has no integrated throat and has the phasing plug directly on its 1.5" output.


Hi POS,

I have seen some of your measurements with 2450-1.5 core:

http://www.audioheritage.org/vbulletin/showthread.php?31657-Truextent-diaphragms-measurements&p=319308&viewfull=1#post319308

but it seems to me that some kind of smoothing is applied, if you have something less-smoothed for 2450-1.5 with H9800.....

Regards
Ivica

[pos]

These measurements are not smoothed, but windowed.
I am not sure I understand what you are looking for, as you will not be using that horn anyway?
Why don't you look for wave tube measurements like the one you can find in the spec sheets?
http://lansingheritage.org/html/jbl/specs/pro-comp/pro-comp.htm

[ivica]

These measurements are not smoothed, but windowed.
I am not sure I understand what you are looking for, as you will not be using that horn anyway?
Why don't you look for wave tube measurements like the one you can find in the spec sheets?
http://lansingheritage.org/html/jbl/specs/pro-comp/pro-comp.htm


Hi POS,

Well, if we are talking about smoothing, or windowing impulse response, averaging data..... it can be said we are talking of some kind of filtering data,
some time it is ' in time domain.', some time in 'frequency domain'
Up to now, my experience showed me, that the more resolution method used the more non-ideal response can be get.
Yes, the influence of near objects can mask the real 'truth' too, but some of them (example horn or box edges are always present).
In post #18, deeps and peaks over 9kHz (red line) I believe has showed , that the diaphragm "entered in modal oscillations",
but with the smoothed (blue line) such behavior is masked, because of 1/3 octave averaging filtered data. Shortening impulse response can show almost the same "results". From the presentation of impulse response
as I have mentioned your measurements in post #22,

http://www.audioheritage.org/vbulletin/showthread.php?31657-Truextent-diaphragms-measurements&p=319308&viewfull=1#post319308

'windowing' that has lasted about 1.5ms (= 0.5/345 ), that would correspond to the filter resolution of about 600~700 Hz...'per filters cells'..

If we are talking about the driver responses loaded with the 'wave tube', the responses of driver 2450-2 looks almost the same as 2451-(1.5), but I am not sure that the same horn would give us the same results for each of them, especially over 10kHz, but under 10kHz (what is my main interest) I would expect quite the same behavior, and just because of that the responses of 2450-1.5 with Ti ribbed diaphragm in H9800 provoke my attention.

Regards
Ivica

[gibber]

The 2452 also has the 4-slit coherent wave phasing plug, albeit a bit shorter. In my tests the 2450SL core outperformed the 2452 one by quite a margin. I don't know how to express that, but the dynamic was more realistic on the 2450. The 2452 sounded more "fake" in this regard. I think is has to do with the "optimized" magnet sturcture of the 2452 (field distortion?...)
I found the 2452 and 2435 both shared this trait...

Given the same throat and horn, and of course the same diaphragm, all the coherent wave 4" CD should measure similar (2446, 2447, 2450, 2450SL/2451...)
One thing that seems to appear though is that the ferrite drivers do not have the dip in the ~3khz range that the neo drivers have. Maybe this is due to the bigger back cap?

( i combined two quotes into one for brevity )

POS, i agree for plane wave tube or longish horn, the 4" drivers likely measure almost the same and sample variation of diaphragms dominates (older frams tend to have two dips at 1.9 and 3.8 kHz).
However on shorter, fast-opening horns (HR9040A in my case), the shorter throats increasingly cause problems. The top end gets slightly more rugged and the range <500 Hz suffers in level. 2431/35 are a disaster on that horn, 2452 i did not try (and don't plan to...). However, the 2431/35 experience got me thinking, especially after seeing the 2432's "throat extension".

For the 4" fram 38mm exit units, I measured length of path from fram mounting to driver exit as follows: 2452: 20mm, 2451/50SL: 31mm, 2447: 47mm. The 2447 and the 50/51 share the phase plug, but the '47 adds what looks like a 16mm straight tube to the driver, possibly causing the 4dB more level @ 400Hz on HR9040A (??) I have yet to get a 16mm spacer for my 2450SL to test the latter hypothesis.

[pos]

2431/35 are a disaster on that horn...That one is a 3" diaphragm, and a totally different phasing plug (3-slit).
I also have many problems with these on short horns (especially the 2332!).


For the 4" fram 38mm exit units, I measured length of path from fram mounting to driver exit as follows: 2452: 20mm, 2451/50SL: 31mm, 2447: 47mm. The 2447 and the 50/51 share the phase plug, but the '47 adds a what looks like a 16mm straight tube to the driver, possibly causing the 4dB more level @ 400Hz on HR9040A. I have yet to get a 16mm spacer for my 2450SL to test the hypothesis.
Interesting, I knew that the 2452 had a shorter phasing plug, but I could not measure any significant difference in on/off axis response.
I thought the 2447 shared exactly the same phasing plug than the 2451/2450SL tho...

What king of difference do you see in your measurements?

[gibber]

That one is a 3" diaphragm, and a totally different phasing plug (3-slit).
I also have many problems with these on short horns (especially the 2332!).


Interesting, I knew that the 2452 had a shorter phasing plug, but I could not measure any significant difference in on/off axis response.
I thought the 2447 shared exactly the same phasing plug than the 2451/2450SL tho...


The 2431/35 have strong response variation with HR9040A on axis. Since the driver is not bad per se (and the horn above reproach), my belief is that the very short phase plug of 2431/35 defeats the CD design of Don Keele's masterpiece. That horn only has a very short 12mm "throat" section after the incoming wave is basically ripped apart to feed the constant directivity expansion. The design seems to work only with drivers that add some integrated throat (as was usual at the time of design). Even the difference in length of "integrated throat" between Altec 288C and 290/288G cores is noticable in listening (and measurement by a few extra dB below 600Hz).

I think it was you who first reported polar response differences between the 4" and 3" fram JBL drivers. If you ask me, this in part could be due to the # of phase plug channels - 3 versus 4 (vs 5 in TAD) but also due to length. The short 243x ad 2452 phase plugs do away with the curved outline introduced as "coherent wave" for the 2446/47/50/51. Not much leeway if you want to curve in such a short design.

You are probably right that the 2447 and 2451/50SL share the phase plug (actually that was my guess in the earlier post). What the 2447 adds is a 16mm long kind of straight "tube" through the front pole plate, whereas the phase plug end in 2451/50SL is flush with the mouting area of the horm.

Update: i just got the 2447 out of the shelf and had another look. The ca 16mm actually is the pole plate thickness, so what JBL does is fit an extra thick pole plate (it is phat) to 2447 which simply carries a straight 38mm bore to conduct the wave a wee bit after it leaves the phase plug. In 2451/50SL, a wave leaving the phase plug has already entered the horn...



What king of difference do you see in your measurements?

For 2447 I measured 4 dB more level around 400Hz and a slightly smoother response in the last octave as compared to 2450SL. Same non-SL fram in both drivers used for the tests.

[pos]

For 2447 I measured 4 dB more level around 400Hz and a slightly smoother response in the last octave as compared to 2450SL. Same non-SL fram in both drivers used for the tests.
+4dB! This is a lot!
Do you think these difference are solely due to that extension, or might the larger back chamber of the 2447 also be at play?

I also did measure some differences in the lower range between the 2452 and 2450SL cores, with the 2452 being slightly more efficient in this range (but I did not test it down to 400Hz...). The 2452 has the smallest back chambers of all these 4" drivers.


The 2450/2450SL/2451 drivers also appear to all have a small dip around 3khz, which is not present in the 2446 (if you look at the planewave tube measurements). I also have this small dip in many of my measurements, and that would be interesting to see if enlarging the back chamber could alter this...

[gibber]

+4dB! This is a lot!
Do you think these difference are solely due to that extension, or might the larger back chamber of the 2447 also be at play?

I also did measure some differences in the lower range between the 2452 and 2450SL cores, with the 2452 being slightly more efficient in this range (but I did not test it down to 400Hz...). The 2452 has the smallest back chambers of all these 4" drivers.


The 2450/2450SL/2451 drivers also appear to all have a small dip around 3khz, which is not present in the 2446 (if you look at the planewave tube measurements). I also have this small dip in many of my measurements, and that would be interesting to see if enlarging the back chamber could alter this...


Alright, it is a little late for this now, but tomorrow i will do the following: get an alu fram (so neither 2447 nor 50SL are on home ground), remove the 8 stand-offs and plug it into 2447 with own back cap and with a 2450 back cap. sounds like a fun thing to do ... ever since the 435be vs 2435 discussion years ago i have wondered how much the back cap contributes and how much the aquaplas on the dome.

57304

57305


The back cap seems to have no influence in this case (both caps actually offer quite generous volume compared to what's available in 2435HPL).


Btw, easy to do as this picture shows - i glued the rubber gasket to the 2450 cap using Uhu Kontakt (the German sniffing community's perennial favourite) and the screws just fit...

57306


Here's the pictures that show the ca 16mm of throat added in 2447, whereas 2450/2451 (and 2452/2453, which i don't own) are throatless drivers

57307

57308

Good sound, i only ran this mono through the HR9040A used for the measurement, but it sure sounds more lively than the TI and TI-Aquaplas. WIll have to check this out in stereo. You need to sacrifice the 8 pcs of stand-off on the alu done to try this out, as otherwise there's a large air flow between phase plug and back volume.

[ivica]

Alright, it is a little late for this now, but tomorrow i will do the following: get an alu fram (so neither 2447 nor 50SL are on home ground), remove the 8 stand-offs and plug it into 2447 with own back cap and with a 2450 back cap. sounds like a fun thing to do ... ever since the 435be vs 2435 discussion years ago i have wondered how much the back cap contributes and how much the aquaplas on the dome.



The back cap seems to have no influence in this case (both caps actually offer quite generous volume compared to what's available in 2435HPL).


Btw, easy to do as this picture shows - i glued the rubber gasket to the 2450 cap using Uhu Kontakt (the German sniffing community's perennial favourite) and the screws just fit...

Here's the pictures that show the ca 16mm of throat added in 2447, whereas 2450/2451 (and 2452/2453, which i don't own) are throatless drivers


Good sound, i only ran this mono through the HR9040A used for the measurement, but it sure sounds more lively than the TI and TI-Aquaplas. WIll have to check this out in stereo. You need to sacrifice the 8 pcs of stand-off on the alu done to try this out, as otherwise there's a large air flow between phase plug and back volume.

Hi Gibber,
Nice and instructive work. Do you have less-smoothed response (say 1/24), even that aluminum diaphragm (in 2441 standard D16r2441, with 'diamond'-metal suspension) in your experiment used, is may be good damping itself.

Mentioned loss around 3kHz can be seen on Mr. Keele original work while presenting HR9040 horn).
For me, it is interesting that you heave no beep notch behavior about 10kH, that is visible on the Mr. Keele measurements ( I believe that is "pipe effect loss" due to its 1.4~1.5-inch dimension). At that time most of the drivers used were with built-in short horn ( mentioned 2441 has about 3 inch long horn 'converting' 1.5inch to 2 inch driver)

[pos]

Very interesting, thank you Gibber!
You should try truextent diaphragms if you can: these are a real pleasure to measure and listen to :)

[ivica]

Very interesting, thank you Gibber!
You should try truextent diaphragms if you can: these are a real pleasure to measure and listen to :)
Hi Pos,
In comparison aluminum D16R2441 and truextent-Be, I understand that you prefer Be diaphragm, is it??
Have you tried, Radian 4" aluminum to compare with others?
My Regards
Ivica

[pos]

Hi Ivica

No, sorry if my comment could me misleading: I did not compare the Be to any Al diaphragm directly: I only compared side by side the tuextent Be, JBL Ti ribbed, and JBL Ti SL...

Here is a comparison between the truextent and the radian:
http://materion.com/~/media/Files/PDFs/Electrofusion/TTB001_SSX_Vs_Radian.pdf

In my comparisons I found the ringing of the ribbed Ti in the UHF (due to breakups) quite audible (once you put your finger on it).
Looking at these measurements it looks like the Radian also suffer from this type of breakups/ringings...

[4313B]

The back cap seems to have no influence in this case (both caps actually offer quite generous volume compared to what's available in 2435HPL).
Impedance curves?

[gibber]

Impedance curves?


The B&K stuff i use to run my acoustic measurements from Arta cost a ton, but sorry, don't have any hardware to do imp plots.
I do Q measurements on cone drivers by hand, but that's just three points on a single impedance peak. Arta doesn't seem to offer imp plots

Your recommendation?

[gibber]

Very interesting, thank you Gibber!
You should try truextent diaphragms if you can: these are a real pleasure to measure and listen to :)

If you have some for me to try, i am happy to give them a go :)
According to rumour i heard, they're no longer available after loss of their Be foil supplier. Rumour true or not? -- definitely haven't seen them on offer any more for a while. In any case i was not super-impressed with the Be drivers i have run in my setup so far (JA801 with modded Jabo, TD4002(Z)/2435/TD4001 on various horns), so i missed out on Truextent volontarily when they were still easy to get.

[gibber]

Nice and instructive work. Do you have less-smoothed response (say 1/24), even that aluminum diaphragm (in 2441 standard D16r2441, with 'diamond'-metal suspension) in your experiment used, is may be good damping itself.

Here you go, Ivica - i doubled resolution of the pic also

57311

==> this is on the same HR9040A horn, and the periodic behaviour (every 190Hz) i attribute to the distance either between horn and mic or more likely, distance between (upwards aiming) horn and my 2.5m heigh ceiling. The driver is placed on the floor during measurement. The arrangement allows me to put horns on drivers fast, while the horn mass in combination with the gasket most drivers have still provide good if not perfect sealing ...


Mentioned loss around 3kHz can be seen on Mr. Keele original work while presenting HR9040 horn).
For me, it is interesting that you heave no beep notch behavior about 10kHz, that is visible on the Mr. Keele measurements


The HR9040 is a slot radiator type horn with 33mm entrance height that is squeezed down to just 18mm vertical size at the point where the slot opens into the radial section.
The HR9040A opens up from 33mm at throat entrance to 38mm vertical where the radial section begins. 38mm is more than double of 18mm.
Both horns expand towards the mouth following almost exactly the same horizontal formula.

As a result, we have two very different horns, even if the most visible part of them (the super-large mouth) is exactly the same and the horizontal shaping is almost the same. Because of slower area expansion in HR9040, the lower cutoff strongly differs, and the "slot" of HR9040 (Keele used it in the JBL bi-radials again, but turned vertical) will surely make for a very different high frequency behaviour.

Not sure the dip in your post of Keele's HR9040 plot is due to either "pipe effect loss" or diffraction, but it strikes me that speed of sound divided by 0.018m slot size is 19kHz, and the response anomaly is at half that frequency.

[Mr. Widget]

If you have some for me to try, i am happy to give them a go :)
According to rumour i heard, they're no longer available after loss of their Be foil supplier. Rumour true or not? -- definitely haven't seen them on offer any more for a while. In any case i was not super-impressed with the Be drivers i have run in my setup so far (JA801 with modded Jabo, TD4002(Z)/2435/TD4001 on various horns), so i missed out on Truextent volontarily when they were still easy to get.Truextent couldn't have lost their supplier... they are one of the few companies in the world that actually mines Be ore. ;)

In any event they are back in production and available.


Widget

[ivica]

Here you go, Ivica - i doubled resolution of the pic also

==> this is on the same HR9040A horn, and the periodic behavior (every 190Hz) i attribute to the distance either between horn and mic or more likely, distance between (upwards aiming) horn and my 2.5m height ceiling. The driver is placed on the floor during measurement. The arrangement allows me to put horns on drivers fast, while the horn mass in combination with the gasket most drivers have still provide good if not perfect sealing ...

The HR9040 is a slot radiator type horn with 33mm entrance height that is squeezed down to just 18mm vertical size at the point where the slot opens into the radial section.
The HR9040A opens up from 33mm at throat entrance to 38mm vertical where the radial section begins. 38mm is more than double of 18mm.
Both horns expand towards the mouth following almost exactly the same horizontal formula.

As a result, we have two very different horns, even if the most visible part of them (the super-large mouth) is exactly the same and the horizontal shaping is almost the same. Because of slower area expansion in HR9040, the lower cutoff strongly differs, and the "slot" of HR9040 (Keele used it in the JBL bi-radials again, but turned vertical) will surely make for a very different high frequency behaviour.

Not sure the dip in your post of Keele's HR9040 plot is due to either "pipe effect loss" or diffraction, but it strikes me that speed of sound divided by 0.018m slot size is 19kHz, and the response anomaly is at half that frequency.

Hello Gibber,

Thanks for the explanation the differences between HR9040 to HR9040A horns, and the measurements 2447 driver ( I understand with D16R2441 ALUMINUM dome) in HR9040A horn.

My expectation of 9kHz deep notch (loss) in Mr. Keele measurements (that are definitely not present in your precise work 1/24 smoothing). I can understand that in such way: if the sound is passing 'between" two parallel plates, that are D apart, then the first notch would be expected around f= Co/(2*D), so for 18mm=0.018m expectable f[18]= 9580Hz, and if D=38mm, f[38]=4540Hz, but in your measurements none of the mentioned frequencies are visible as notches.

Regards
Ivica

[gibber]

Truextent couldn't have lost their supplier... they are one of the few companies in the world that actually mines Be ore. ;)

In any event they are back in production and available.


Widget


Widget

thanks for the info,
do Trueextent mine Be and do the foils or just mine Be and have Be end user products ?
That would be interesting to know, as i heard from a French speaker designer there are only two Be foil companies ww, so Truextent might have the whole value chain in-house.

Sad i never saw the offer Altec large format drop-in replacements. Talking about drop-in of diaphragms (and this may or may not warrant another thread), does anyone know the meaning of JBL's magic numbers ? Here's a few shots:


573235732457325

573265732557327

The "+1" fram came from the "-3" driver which looked like never opened (but i bought it 2nd hand...)

[gibber]

if the sound is passing 'between" two parallel plates, that are D apart, then the first notch would be expected around f= Co/(2*D), so for 18mm=0.018m expectable f[18]= 9580Hz, and if D=38mm, f[38]=4540Hz, but in your measurements none of the mentioned frequencies are visible as notches.

Regards
Ivica

Hi Ivica,

the HR9040 monotonously increases area like all good horns do, but being square in area, it can do so while severly tapering vertical size from 33mm to 18mm. I don't have HR9040, so the 9.5kHz notch cannot possibly be in my measurement screenshots. For more info, see posts ab0ez and i did elsehere. The 1975 Keele paper you quoted has a drawing of vertical expansion. Horizontal expansion of HR9040 and HR9040A are the same except for the first 1cm into the horn. That first horizontal 1cm is very different, though. It's also printed in the 1975 paper.

The HR9040A does not taper the incoming 33mm at all, but rather opens up gradually. I can see your reasoning for an artifact at 4.5 KHz. Not sure we should expect a pronounced diffraction effect, if the horn opens up. I have maybe 50 more measurements of HR9040A with about 15 different drivers. I just went through them and saw there is a tendency (visible with about two thirds of the drivers) to have a reduced amplitude by about 3dB @ 4kHz, not 4.5 kHz. This drop is not narrow as in the plot by Keele you attached, but i will re-do some of these plots in 24th oct resolution to see whether your prediction is right. I will upload if i find something useful.

Ralph

[ivica]

Hi Ivica,

the HR9040 monotonously increases area like all good horns do, but being square in area, it can do so while severly tapering vertical size from 33mm to 18mm. I don't have HR9040, so the 9.5kHz notch cannot possibly be in my measurement screenshots. For more info, see posts ab0ez and i did elsehere. The 1975 Keele paper you quoted has a drawing of vertical expansion. Horizontal expansion of HR9040 and HR9040A are the same except for the first 1cm into the horn. That first horizontal 1cm is very different, though. It's also printed in the 1975 paper.

The HR9040A does not taper the incoming 33mm at all, but rather opens up gradually. I can see your reasoning for an artifact at 4.5 KHz. Not sure we should expect a pronounced diffraction effect, if the horn opens up. I have maybe 50 more measurements of HR9040A with about 15 different drivers. I just went through them and saw there is a tendency (visible with about two thirds of the drivers) to have a reduced amplitude by about 3dB @ 4kHz, not 4.5 kHz. This drop is not narrow as in the plot by Keele you attached, but i will re-do some of these plots in 24th oct resolution to see whether your prediction is right. I will upload if i find something useful.

Ralph

Hi Gibber,

Looking at the mentioned Mr. Keele presentation HR9040 it seems to me that in vertical plane there is almost about 10cm ( 4-inch) part of almost parallel surfaces, but as I have understood you, on HR9040A there is "almost" the same 'situation' but 38mm (or some mean value form 33mm to 38mm) not 18mm distance. But there is no any 'intention' of notch-filtering effects, around 4.5kHz.

[gibber]

Hi Ivica, just in case you own HR9040, would you say that the drawing you attached is correct? I ask because from photos i saw i seem to remember the vertical tapering does not happen at the horn throat only, but is more gradual. Maybe i'm mistaken, it's been a year since. Here's one of the measurements where the driver showed a 4kHz 3 dB droop on HR9ß40A -- redone for you in 24th oct resolution. The meas conditions are not the same as they were when i did the 6th oct plot last year, but are similar as for the screenshots i posted in this thread before. Maybe ca 2dB are broadly missing in the area around 4.5 kHz. But a strong notch effect is not there. Let's maybe take further discussion of the EV horn into pm, as i believe this thread veers off-topic quite a bit already.

57330

[ivica]

Hi Ivica,
.................Let's maybe take further discussion of the EV horn into pm, as i believe this thread veers off-topic quite a bit already.

[]
Hi Gibber,
Many thanks for your precise work, I have sent you PM.
Regards
Ivica

[Mr. Widget]

Widgetthanks for the info,do Trueextent mine Be and do the foils or just mine Be and have Be end user products ?That would be interesting to know, as i heard from a French speaker designer there are only two Be foil companies ww, so Truextent might have the whole value chain in-house.Truextent is part of a much larger company. Part of the company mines the ore in Utah, part of the company creates the foil in the midwest, and the part here in California creates the Truextent diaphragms that we have been talking about as well as domes and foils for other loudspeaker companies.


Widget

[marco_gea]

Here's the pictures that show the ca 16mm of throat added in 2447, whereas 2450/2451 (and 2452/2453, which i don't own) are throatless drivers

57307

57308


I know this is a very old thread... but I thought I'd chime in on this difference in the internal throats of the 2447 vs 245x drivers.

Actually, if you ideally add to both drivers the 63mm-long "snout" that the 2450 has to lead to the classic 2" (49mm) exit, you get the following internal flare rates (based on the equivalent exponential expansion): 180Hz for the 2447, and 225Hz for the 245x.

This is interesting, as it appears that the 2447 + "snout" = 2446, which shares the same flare rate with all previous historic 244x drivers back to the original Lansing 375.

Instead, the 245x drivers share a different (newer) design that happens to have the exact same internal flare as the TAD TD-4001 and TD-4002. This can be easily verified by doing the calculations on the removable "snout" of the TD-4002, which is 64.5mm long and goes from a starting diameter of 39mm to an exit diameter of 50.8mm.

I thought this "coincidence" was cool, and decided to post this for future reference.

Cheers,
Marco