PDA

View Full Version : Nomograph for -6dB point dispersion angle Frequency vs. Diameter



Robh3606
09-02-2005, 09:33 AM
Though some of you might find this useful.


Rob:)

Zilch
09-02-2005, 11:27 AM
15" driver is 100° nominal at 1 kHz.

That'd be 4430's touted beamwidth match between LF and HF. :)

[More like 110°, actually, for -6 dB.... ]

Dr.db
03-27-2018, 02:06 PM
If 15" driver is aprox. 100° at 1khz, why do most 3-way speakers cross significantly lower?

For example:
S9800 at 750hz
S9900 at 850hz
TAD 2402 650hz

speakerdave
03-30-2018, 02:45 PM
My opinion:

Because in a cone driver meant to reproduce low bass other factors must be considered, such as cone mass, suspension slackness, woofer response tilt, and range of pistonic behavior. Since the issue when the driver is in effect too small to match the treble beam width is power distribution in the room it only makes sense to obsess about it if the whole room is going to be tested and treated just as carefully. Since home listening rooms almost never are, and studios can be tuned as needed, the speaker designer is free to make good bass and lower the crossover if necessary.

A mismatch of cone and horn beam width is more critical when the cone is too big for the crossover frequency and its beam width is too narrow, because the resulting sound can be very harsh. Note how the horn pattern was narrowed in the Altec Model 19.

Besides that, I think there are still plenty of people who believe on-axis response is the truly crucial matter.

Flaesh
03-30-2018, 11:39 PM
One more:

Dr.db
03-31-2018, 02:51 PM
@Dave:
Your explanation makes sense, that dispersion match is not the only factor affecting the choice of crossover-frequency...

@Flaesh:
Thanks for that chart.
If i´m reading this correctly, 90° dispersion of a 15" cone would be actually higher, more like 1,3khz. That seems way to high as a crossover for a 15" cone.
A 12" or 10" reaches 90° dispersion even much higher...
So if you are aiming a crossover of ~800hz to a 90° horn, a 12" or 10" would be even much worse than a 15" regarding dispersion-match, correct?


I´m wondering if this chart explains the poor imaging-abilities of the older monitors such as 4343 or 4345.
They had a 10" crossed at 1,2khz to a horn with 80° dispersion.
At 1,2khz the 10" cone has a dispersion of aprox. 150°.
Not matching at all.

A system like the S9900 is said to image a lot better, maybe this is because the 15" cone matches better with the 90° horn...

Robh3606
03-31-2018, 03:56 PM
A system like the S9900 is said to image a lot better, maybe this is because the 15" cone matches better with the 90° horn...

It's the type of horn and having a reasonably smooth transition where there are no abrupt changes in the DI through the crossover region. So yes you do want a close match at the crossover point. But the horn ultimately controls and contributes the most to the imaging. Check out the 4430 AES paper on this site. It's linked in the post. Virtually all of the modern JBL horn systems are designed using Constant Directivity horns and have their roots in the 4430 design. Take a look at the DI plots for the 9900 also posted here.

http://www.audioheritage.org/vbulletin/showthread.php?7852-Improvements-in-Monitor-Loudspeaker-Systems



Rob:)

ivica
04-01-2018, 02:16 AM
@Dave:
Your explanation makes sense, that dispersion match is not the only factor affecting the choice of crossover-frequency...

@Flaesh:
Thanks for that chart.
If i´m reading this correctly, 90° dispersion of a 15" cone would be actually higher, more like 1,3khz. That seems way to high as a crossover for a 15" cone.
A 12" or 10" reaches 90° dispersion even much higher...
So if you are aiming a crossover of ~800hz to a 90° horn, a 12" or 10" would be even much worse than a 15" regarding dispersion-match, correct?


I´m wondering if this chart explains the poor imaging-abilities of the older monitors such as 4343 or 4345.
They had a 10" crossed at 1,2khz to a horn with 80° dispersion.
At 1,2khz the 10" cone has a dispersion of aprox. 150°.
Not matching at all.

A system like the S9900 is said to image a lot better, maybe this is because the 15" cone matches better with the 90° horn...

Hi Dr.dB,

YOu have to read horn dispersion characteristics with greater attention, as most of them have much wider dispersion at lower frequencies, then their nominal specified data. Only very large mouth horns can be specified dispersion on the lower frequencies.

for example (2425 with 2307 and 2308 lenses would have over 140 deg dispersion round 1kHz, I can imagine that 2440/.41/.45/.46/.50 with 2311 and 2308 would behave almost identically round 1kHz or less, if using -6dB level as 'cut-off dispersion level' ):

http://www.jblpro.com/pub/obsolete/Acoustic_Lens_Family1.pdf
(http://www.jblpro.com/pub/obsolete/Acoustic_Lens_Family1.pdf)
while using 2360A horn, the dispersion round 1kHz would be much more controlled relative to the specified data.

regards,
ivica

Robh3606
04-01-2018, 08:41 AM
Hello Ivica

Here is an actual plot from the referenced AES paper. They have the Urei Coax and 511 Altec horn in there as well.

Rob:)

Dr.db
04-06-2018, 01:39 PM
@Ivica:
You are right with these lenses, I can see that in your posted graphs. Thanks.
So I was wrong as the lenses should match the dispersion of a 10" actually quiet well!
So the poor imaging compared to more modern designs must be in the horn itself as Rob mentioned.

@Rob:
I´ve read that paper, thank you.
So the key to good imaging is a Constant-Directivity Horn and a directivity match with the woofer below.
But I would need a equalization e.g. CD-compensation... If you don´t use a M553 crossover, how can this be applied with a usual analog-active-crossover like marchand ?

What about the Yuichi A290 for example?
This is not a CD-horn, right? But do the vanes help control the directiviy-index?
So how does it image?
Does it have 90° horizontal dispersion at ~800hz or is it much wider in realitiy?

Robh3606
04-07-2018, 08:06 AM
But I would need a equalization e.g. CD-compensation... If you don´t use a M553 crossover, how can this be applied with a usual analog-active-crossover like marchand ?


You would need to measure the system with the horn mounted to the baffle to see what the response is. Then you would just figure out a voltage drive that is the mirror of the measurement to flatten things out. Canned curves are only aproximations and wont get you the best possible response. The easiest way is just to use a passive network if you are going analog. But that means you need something like LEAP to do the network. If you go digital it's much easier.


Rob:)

Ian Mackenzie
04-07-2018, 01:19 PM
The topic of imaging is more than just the horn geometry

In theory above 1000 hertz the transducer should ideally be a point source at a 3 dimensional point in space.

The M2 horn and D2 driver is claimed to approach the ideal

That involves a lot of big ask technical challenges when it comes to implementing with alternative drivers by diy people who are inclined to disagree on what works and what doesn’t work.

For example will a single Be diaphragm compression have sufficient high frequency extension? On paper it looks like it does but subjectively there are opposing views. So some end up adding a uhf horn to extend the highs. But then it’s no longer a 3 D point source. So they grab a dsp crossover and add delays in an attempt to solve that problem. This adds a lot of complexity. Many audiophiles with analogue sources and digital sources argue even the very best out of the box dsp crossovers are not completely transparent and they buy soa outboard ADC and DAC. That requires significant financial resources.

The difficulty for Jbl diy fans these days is taking a concept based on a Jbl engineered design and then implementing it at home with alternative horns and drivers because Jbl no longer supply the drivers and horns directly.

At a more practical level a pure constant CD loudspeaker is claimed to work best when placed in the corners

But depending on your room dimensions and how the room is arranged this might not be appropriate.

Throw horns out of the conversation for a moment and there are numerous multiway direct radiator loudspeakers that are claimed to have outstanding imaging properties. But this comes at a price according to horn loudspeaker aficionados.

However, active loudspeaker technology is starting to gain momentum and the capability gap between conventional direct radiator systems and horn systems is getting smaller.

The point is you can’t look at one loudspeaker performance parameter in isolation. Focussing on one aspect alone could come at a huge penalty overall or involve significant expense. So it’s about getting the balance right with a set of complementary compromises.

In the diy space that will keep you busy.

Dr.db
04-07-2018, 02:48 PM
Thanks a lot, both of you.

How would you judge the A290 Yuichi´s imaging abilities?

Probably the yuichi matches the best with a 15" cone somewhere around +-800hz, right?
Which crossovers would you recommend between a 10" cone (2123h) and the yuichi ?

Btw, how important is a good cone-horn match in reality ?

Earl K
04-07-2018, 05:27 PM
db,

Try to absorb the significance of the following quote ( regarding imaging ).


I'm Back! (if only for a moment)


Congratulations to those who have received their 4367's. There are a ton of misconceptions regarding the speakers. i can comment through the DV run after which I was no longer involved. The speaker is essentially similar to a 4365 in overall sound performance. There are some voicing differences partially due to my desires and partially as a result of the hardware used. The 2216Nd has the low TCR wire and is very kicky and dynamic as a result. The 2430 needs to be used an octave below where it is happy (as in the M2) and has an increased amount of 2nd harmonic distortion between 750 Hz and 1500 Hz as a result. It is not horrible, but it is about 10 dB higher than the 476Mg in that range and up to 20 dB higher than the 476Be over the same octave. The increased second comes from the 2430 being a ring radiator with no suspension. It is just difficult for it to move below 1500 Hz. The Mg diaphragm is twice as thick as the Be diaphragm so it resists motion in that range to a degree. The Be has the most freedom to move in the octave at and above crossover so it is the best of the three in that regard. With that said, the 2430 has a very detailed and musical sound and is a fine driver. It is capable of much higher output levels than any of the 476 family and is therefore well suited to Sound Reinforcement and Studio use. There is no silver bullet. With good things come bad things and one has to look at the total set of compromises. The 2430 is also about 1/3 the cost of the 476 precious metal drivers.

The design goal of the 4367 was to equal or surpass the performance of the 4365 in a smaller enclosure and for 1/3 less money! Done and Done. The system will thrive on LF EQ. There is plenty of headroom in the woofer so 4 - 6 dB of boost around 32 Hz will really spice up the mix.

The system has very nice imaging but it cannot touch any Array in that area because the horn orientation is the wrong direction and the horn is in a wide boxy enclosure. The M2 does better in this regard because the horn is symmetrical in pattern and designed to have a significantly wider coverage pattern. That just can't be done in such a compact horn as those in the 4367 and 4365. BTW, those two horns behave very similarly with neither having a major edge on the other in measured performance. They are not, however interchangeable physically and would require different EQ.

Also keep in mind that the M2 is full active. It has separate amplifiers, electronic and digital crossovers and a lot of EQ bands. All kinds of things large and small can be fixed with that kind of horsepower. The pure passive systems can only have a little shaping and perhaps a few low Q correction filters. In spite of that, there is often a musicality to passive systems that purely electronic ones just can't match.

With regards to charge coupling, the voltage needed to bias the cap pair is pulled from the input through a large resistor (to limit current) and a low leakage diode. The caps in the system are of very high quality so they retain the charge for a while. The beauty of this system not only eliminates the need for the battery and housing/mounting mechanisms but it tracks the signal level. At low levels, very little bias is necessary to do the job so things are fine. At really high levels, much more bias is necessary but the generation level goes up so things are fine. With a fixed battery voltage, it is possible to run out of bias at high levels. The diode method is a good improvement and in my opinion, sounds better than battery biasing. So suffer through the first 1 or 2 seconds of your first playing and then forget about it.

The 4367 is a good system for those who like the large Monitor format. It measures well, sounds detailed and musical but is lean in the bottom octave as are all of the post 1985 or so "Japan" product. The speed and excess excursion capacity of these systems makes them good candidates for EQ, or for the proper use of a subwoofer, say below 40 Hz. They do have a "live music" sound that is most difficult to achieve purely with direct radiators. If you are looking for an Audiophile loudspeaker with 3-dimensional imaging, a warm mellow midrange and no dynamics at all, look elsewhere. Without using the words Distilled Water you might look at another Harman brand if you are seeking elegant, luscious elevator music.

I used some concrete blocks to elevate the 4367's when I did Demo's at the factory. 6" to 8" height really helps. The systems are too short for typical American use. The trick is to get something that doesn't rock or tip. You will eat the bass kick immediately if any enclosure movement is allowed. The woofer kicks really hard and if any energy is consumed moving (rocking) the enclosure due to it not being properly coupled to the floor, you will lose bass kick and impact. The proper use and need for spikes (or equivalent) is not BS. It is based in sound theory and is easily demonstrable.


The Array horn has a horizontal coverage something like the following pic from POS ( though since the horn is taller, the polar patterns maintain their pattern spreads to a lower frequency ):

http://www.audioheritage.org/photopost/data//500/H9800_2435H_vertical.png

If you want to experience good imaging, one wants to involve one's ears in an audible 6db down-point, that also has nice constant directivity delineation, as seen in the pic ( & is maintained from at least 1K to 10K, with nicely spaced polar lines ).

You'd do well at your young age to discover the paradoxes that lie behind imaging ( which so far have been ignored in this thread ).

Build up something like what RobH shows in his moniker ( any le14 variant will do, as will almost any Altec 811 sized Radial horn turned vertically ).

If you do this, you will eventually forget about the industry wide SR dogma ( of the so-called necessity ) to match polar patterns at crossover.

JBL's best ( horn based ) imaging champ doesn't do that ( it actually beams much like a flashlight > or StarWars "Light-Saber" in the HF ).

:)

Here's the Timbers Array Thread (http://www.audioheritage.org/vbulletin/showthread.php?9951-Greg-Timbers-amp-DIY) for a bit more background on the subject.

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=13892&stc=1&d=1142807250

ivica
04-08-2018, 05:19 AM
Thanks a lot, both of you.

How would you judge the A290 Yuichi´s imaging abilities?

Probably the yuichi matches the best with a 15" cone somewhere around +-800hz, right?
Which crossovers would you recommend between a 10" cone (2123h) and the yuichi ?

Btw, how important is a good cone-horn match in reality ?

Hi Dr.db,

I have no personal experience with the mentioned A290 Yuichi horn, but almost any owner , here on the AHF Forum said that it is very amusing sound reproducer.
As it has fins in the horn throat, so some kind of internal reflections must have been present, but I have no idea about that influence to the 'imaging'. As known to me almost ANY wide dispersion horn has some kind of internal reflections including M2, DD67/66K, H9800, H9500 , 2370/80......as it is difficult to be 'protected' from the horn beaming for the frequencies over 10kHz. Such internal reflections and diffraction would produce the 'movements' of the sound source depending of the frequencies, as diffraction 'lines' behave as another source, but not for all frequencies, mainly for the higher frequencies.

I believe that You will be satisfied with A290 (with the fins), without fins, I would expect beaming over 10kHz.
regards
ivica

Ian Mackenzie
04-08-2018, 12:00 PM
Dr DB

Check the A290 fin and no fin response on the website

Just email Yuchi

He is a nice guy and l have shared communication

The George Ausberger studio monitors use fin horns

They are an industry bench mark

The devil as always with audio is in the details

I would also talk to some seos users

IMHO imaging is as much about diffraction-time domain related aberrations as it is about dispersion

It’s worth reading Earl Geddes website on the topic

Dr.db
04-10-2018, 10:28 AM
@Earl:
The "vertical" arrangement of the midhorn seems to work great for imaging.
I believe I have to read the mentioned thread from GT to understand this...

As seen in your posted graph, a linear frequency response under angle seems to be a major factor, right?

@ivica:
The fins seem to affect the high-frequencies more than the mid-frequencies and this results in a different dispersion over frequency. Did I get you right?

@Ian:
The yuichi-horn without the fins seems to have a rising freq-response on axis. The yuichi-horn with fins is rather flat on axis. But what does is tell about the imaging ability?

Ian Mackenzie
04-10-2018, 11:25 PM
My recommendation is to contact Yuichi via email and ask your questions

ivica
04-11-2018, 10:37 AM
@Earl:
The "vertical" arrangement of the midhorn seems to work great for imaging.
I believe I have to read the mentioned thread from GT to understand this...

As seen in your posted graph, a linear frequency response under angle seems to be a major factor, right?

@ivica:
The fins seem to affect the high-frequencies more than the mid-frequencies and this results in a different dispersion over frequency. Did I get you right?

@Ian:
The yuichi-horn without the fins seems to have a rising freq-response on axis. The yuichi-horn with fins is rather flat on axis. But what does is tell about the imaging ability?

Hi Dr.db,

Neglecting some dispersion possible irregularities on the mid frequencies that can be seen on some horns, Yuichi A290 horn fins have main role to enhance horizontal horn dispersion over 10kHz. As Ian said, without fins it can be seen rising F/R ON AXIS, but off axis the dispersion would decline (due to the piston radiation pattern characteristic), so I would suggest to use A290 (original designed) with the fins.
Introducing a kind of diffraction slot in the horn throat (2328) with 2350 horn an improvement in the horizontal horn dispersion can be get without using "fin type solution"

http://www.jblpro.com/pub/obsolete/Radial_Horns1.pdf

almost the same used in DD66/67K , H9500, H9800....

but 'fins type solutions' is used in the A290 solution, TAD TH-4001 solution or mentioned Augspurger solution.

http://www.geocities.jp/arai401204/Horn/A290/A290.html
(http://www.geocities.jp/arai401204/Horn/A290/A290.html)http://www.geocities.jp/arai401204/Horn/HornDesign/Experiment.html
http://quarkconcept.online.fr/tad2401.htm
http://audio-database.com/PIONEER-EXCLUSIVE/speaker/model2401twin-e.html
http://www.audioheritage.org/vbulletin/showthread.php?19390-Augspurger-Horn-Pictures&p=195146&viewfull=1#post195146

Interestingly on the ebay.com may be helpful:
https://www.ebay.com/itm/Play-wood-horn-TAD-TH-4001-vintage-speaker-drivers-horns-perfect-Clone-/302444249871



regards
ivica

Dr.db
04-11-2018, 01:25 PM
@Ian:
I have tried to email yuichi several months ago, but he hasn´t responded yet.

@Ivica:
Thanks for all the links!

.

BTW, does the yuichi-a290 has a 90° horizontal dispersion like the original TAD ?

If I´d use a 10" midwoofer up to the yuichi, which crossover would you recommend?