Handmade Ersatz M9500 Speakers [Archive]

linear

09-25-2006, 07:08 PM

Thanks for your interest.

First the bad news. I don’t have any pics of the construction phase (which took about a year). As well, there is no “physical” JBL content in these speakers, but there is a lot of JBL “intellectual” content.

The cabs are constructed from 1” thick MDF and were covered with red oak veneer. The top box is 5 cu ft (25”x25”x19”) and the bottom is10 cu ft (43”x25”x19”). Like the JBL M9500, they are tuned to 35 Hz and 28Hz respectively. They are necessarily bigger in volume than the ‘9500s (2.8 cu ft & 4.1 cu ft), due to the non-JBL drivers (sorry!) used, but as a result, the response curve matches that of the 9500 almost perfectly.

I turned the horns and the ports myself, out of poplar wood, on a $200 lathe! The round horn is exponential taper and as close a copy to the Zingali units as I could manage on my crude equipment.

The bass drivers are Usher 15HM (which are used in the Usher D2 speaker) and the compression drivers are B&C DE500 1” units (used by Zingali). (The Usher D2 has two identical 10 cu ft cabs, each tuned to about 28 Hz. So, my “stagger tuned” design is pure M9500.)

The system is bi-amped, with the 15HMs (in parallel) and the DE500 driven directly, just like the JBL DMS-1. I designed and built a custom active analog crossover unit (similar to the JBL DX-1) that uses high-speed buffers (no opamps!) and has the filter components on small daughter boards. I am presently using a 4th order Linkwitz-Riley alignment crossing over at 800 Hz. I also have 4th and 3rd order Bessel daughter boards built, but I have not tried them yet. No delay was required, as the horns and woofers measured in phase at 800 Hz. This surprised me somewhat, due to the fact that the DE500 voice coil is about 6” behind that of the 15HM. However, this is also the geometry of the DMS-1, and JBL spec a zero delay for that cab, so I guess all is well.

How do they sound? Well, I guess the fact that my wife is actually happy to have these hybrid monsters in our living room speaks for itself!

Linear

louped garouv

09-25-2006, 07:23 PM

absolutely stunning... :applaud:

linear

09-25-2006, 08:35 PM

speakerdave

09-25-2006, 08:44 PM

OK, I'm guilty of plagarism, but I only steal the best ideas!

Who knows, maybe there will be an "Usher Heritage Form" 20 years from now, but somehow I doubt it! However, Usher does offer a pretty good driver for the price (In stock for $218 at Parts Express), and they even publish distortion curves for that unit.

The 1400nd and 1500al are beautiful speakers, but even if I could afford the cost, availability is somewhat problematic (to put it mildly). If anyone wants to donate some JBL drivers to the cause ("Project May" for Linear :-), I will be more than happy to do a retrofit!

Linear
You certainly are good-natured about my little sarcasm, which you can see I thought better of and pulled down, apparently just before your post came up. Congratulations on finishing what I'm sure was a demanding project.

David

Guido

09-26-2006, 03:31 AM

OK, I'm guilty of plagarism, but I only steal the best ideas!

Who knows, maybe there will be an "Usher Heritage Form" 20 years from now, but somehow I doubt it! However, Usher does offer a pretty good driver for the price (In stock for $218 at Parts Express), and they even publish distortion curves for that unit.

The 1400nd and 1500al are beautiful speakers, but even if I could afford the cost, availability is somewhat problematic (to put it mildly). If anyone wants to donate some JBL drivers to the cause ("Project May" for Linear :-), I will be more than happy to do a retrofit!

Linear

I also plan to work with these Usher drivers. They use the original 2235H cone BTW.
There was an interesting Project with two Ushers in a german speaker builder magazine. I'll try to find better pics but it's down right on the title here:
http://www.hobbyhifi.de/Aktuell/Aktuelle_Ausgabe/aktuelle_ausgabe.html

What HF compensation do you use?

linear

09-26-2006, 09:39 PM

I also plan to work with these Usher drivers. They use the original 2235H cone BTW.
There was an interesting Project with two Ushers in a german speaker builder magazine. I'll try to find better pics but it's down right on the title here:
http://www.hobbyhifi.de/Aktuell/Aktuelle_Ausgabe/aktuelle_ausgabe.html

What HF compensation do you use?

Guido,

I do not use any HF compensation, like JBL do with the DMS-1. The DE500 is somewhat of a poor man's TAD 4001. It is very flat when compared to the compression driver used in the DMS-1 (JBL 2450). So, I am driving everything flat across the audio spectrum.

It's great to know that the 15HM is "partly a JBL" (please, no flames! :-), even if it is just a 2235H cone. Is Usher just using the same cone supplier that JBL once used? I think I remember reading that it was a German company.

By the way, Guido, if you are really considering buying some 15HMs, they are now on sale for $159 at www.partsexpress.com (http://www.partsexpress.com) (item: 296-630). Parts Express will ship internationally (to Germany) by UPS, and they will accept PayPal for payment. Back in 2003, I saw a German price of 350 Euro quoted in "Hobby HiFi" for the 15HM, so this might be a good deal even with the international shipping. It's almost as good a deal as the PE sale on the JBL 1500SUB sub-woofers a year or two ago, and those units sure sold out fast. Luckily, I got two of them.

Speaking of "Hobby HiFi", I would love to see some more of their 15HM article, if you have a chance.

Linear

Mr. Widget

09-26-2006, 09:52 PM

...even if it is just a 2235H cone.Even if it is the same cone... Do they use a mass ring?, Since the surround is a fabric surround, even if the cone, mass ring, voice-coil, spider, and motor were the same... it would still be something other than a 2235. It may be an excellent woofer, I have no idea, but it certainly isn't a 2235.

From what I can tell, Usher is really good at making drivers that look like other people's drivers... I am not so sure they sound like other people's drivers. (I am thinking about their various tweeters and kevlar mids etc.)

Widget

jack_bouska

09-27-2006, 12:33 AM

as a result, the response curve matches that of the 9500 almost perfectly. I turned the horns and the ports myself, out of poplar wood, on a $200 lathe! The round horn is exponential taper and as close a copy to the Zingali units as I could manage on my crude equipment.

. I am presently using a 4th order Linkwitz-Riley alignment crossing over at 800 Hz.
Linear

Very nice looking cabinets and horn, job well done!

I think your placement of the ports on the top and bottom of the cabinets is an improvement over the Usher D2 design, which has the drivers on the top and bottom, and the ports inboard, close to the horn. I believe your configuration will provide better MTM imaging, and the staggered distance of the driver centres from floor and ceiling will act to distribute modal response in the listening room, compared to a D2. (Nice to see a cabinet with edge bevels large enough to be acoustically useful)

I also heartily approve of the use of the axially symmetric horn, which appears to be well integrated into the front baffle. These type of round (simple) horns should generate less "horn honk" compared to bi-radial, or standard square mouth old-school horns. (modern finite-element designed waveguides excepted)

The horn you turned appears to be blocked up from planks of Poplar, with the plank at the mouth consisting of a single, full width, 15" plank! (you must have had trouble sourcing that). You also describe the horn contour as exponential, yet the wooden section appears to be too flat faced to be pure exponential, however photo's tend to distort depth perception so I might be mistaken.

Could you post a diagram of the exact contour, I'm curious.

I also note a short black tunnel running from the compression driver to the wooden section. This might form the exponential throat.

The B&C has a recommended crossover of 1.5 kHz (2nd order), and you are crossing over about 1 octave below that, with 4th order, so you might want to be cautious with the volume control. Adding that extra octave of bandwidth on the low side results in somewhere between 3-6db more power dissipation in the voice coil (based on typical music spectral density). The 4th order will have better attenuation below 400 Hz, but it's the extra midrange energy above 800Hz that may potentially cause damage in your alignment.

I would also like to see your measured response curves (on and off axis), as you mentioned that you have not used any response tailoring. The exponential horn contour has strong directivity narrowing with increasing frequency, which will compensate for the mass breakpoint induced HF response roll off of the compression driver. (thus flattening the response)

I suspect you will enjoy these speakers well into the future,
Jack

Guido

09-27-2006, 10:39 AM

This is available through the usher website so I think there are no copyright issues.

Guido

09-27-2006, 10:49 AM

Some data 15HM (2235)

Gap depth 8mm (7,14)
Winding depth 19mm (19)
BL 16,1 (20,8) !!!! There is a new Usher with stronger magnet released
MM 92 (150)
Vas 463l (525)

This is certainly no 2235 which I never said. The newer model with the stronger magnet looks very much like a 2234 but I don't have a scan at hand.

Hoerninger

09-27-2006, 11:27 AM

If we don't discuss JBL, it's difficulty to get them in Germany,
I'm inclined to mention European manufacturers like
18Sound, B&C, BMS, Beyma, Ciare, Craaft, Oberton and others.
They are worth a look.

But I really want to add, not every car is like Mercedes or BMW. ;)
____________
Peter

linear

09-27-2006, 08:56 PM

Very nice looking cabinets and horn, job well done!

....the Usher D2 design......has the drivers on the top and bottom, and the ports inboard, close to the horn.....

......the mouth consisting of a single, full width, 15" plank! (you must have had trouble sourcing that). You also describe the horn contour as exponential, yet the wooden section appears to be too flat faced to be pure exponential, however photo's tend to distort depth perception so I might be mistaken.........

Could you post a diagram of the exact contour, I'm curious.

I also note a short black tunnel.......

The B&C has a recommended crossover of 1.5 kHz (2nd order).....

......that may potentially cause damage in your alignment.

Jack

Jack,

First let me say that your own project, detailed elsewhere on this forum, is very very impressive, with incredible detail and measurements. Any member who hasn't checked it out, should do so! Unfortunately, I won't be able to meet that standard of documentation.

The Usher D2 cabs can be turned either way (there are 3 separate sections to the speaker). The US rep for Usher says that the "photo shoot" for their brochure was set up wrong! (JBL would NEVER make a mistake like that, right?) I've attached a photo of a D2 that is set up "correctly".

Now, regarding my own project:

The front piece of poplar was two 10" boards glued edge-to-edge. The wood part of the horn is 15.5" dia at the mouth and 6" deep. It has a 2" throat. I use a 3" deep, cast black anodized adapter (JBL designed, I think), to attach the B&C DE500, which is a 1" unit. All of this is exponential taper with a total depth of 9" (from 1" throat to 15.5" mouth). Check out the photos that show all this.

B&C spec the DE500 on their ME45 horn, which has a 1" throat with a 11"x4.5" mouth, and 4.9" depth. This cuts off at about 1 kHz, acording to their spec sheet. The set-up I have described above cuts off at a significantly low frequency, due to the large depth and mouth. My calculations and "crude" measurements indicate a number between 600 and 700 Hz. So, I think I'm OK with a 800 Hz crossover, and that is what Zingali use in their loudspeakers.

One major reason that you don't see any "shallow" round horns, like the above design, in commercial production, is the Guiseppe Zingali holds a US patent 6079514 (and other worldwide patents) on any "round horn in which the mouth diameter is larger than the depth". You can check this out my going to www.pat2pdf.org (http://www.pat2pdf.org) and entering the above number. (It's a free service to download any US patent!)

Linear

MatthiasA

09-30-2006, 10:28 AM

WOW !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
which amp do you prefer ?

qxlxp

09-30-2006, 08:28 PM

One major reason that you don't see any "shallow" round horns, like the above design, in commercial production, is the Guiseppe Zingali holds a US patent 6079514 (and other worldwide patents) on any "round horn in which the mouth diameter is larger than the depth".
Linear

i guess in combination with such startling and revolutionary design breakthroughs such as using not simply wood, but knot-free, wood for a horn, a us patent was inevitable :blink:

you have to wonder who they have working at the patent office.

linear

10-01-2006, 07:13 AM

WOW !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
which amp do you prefer ?

One of the main reasons for undertaking this speaker project was that the wood turning and cabinet building tasks were a pleasant diversion from my 36 year electronics engineering career as a linear analog circiut designer. I have specialized in zero global feedback/local feedback configurations, in an attempt to make solid state circuits sound more like tubes.

Therefore, almost all of my electronics is custom built. This includes my active crossover (previously described), and my power amps, which are a local feedback, fully complementary bipolar design. They can output 300 watts into 4 ohms (the two 15HM LF drivers), or 150 watts into 8 ohms (the DE 500 HF transducer). Since the speaker system is very efficient (98 dB), this equipment will produce an SPL level of almost 120 dB in my living room. (This is more than adequate, according to my wife!)

My system presently uses a Wadia 861 CD player (www.wadia.com (http://www.wadia.com)) driving my custom crossover and power amps. There is no preamp or other equalization used. (I also have a Revox reel to reel and a Thorens turntable with Stax cartridge, now seldom used.)

Although I did not build the 861, I did design and patent the analog circuitry of this player, and licensed it to Wadia. They call it their "swift current" block. If anyone's interested they can check this out by entering my patent 4983930 into www.pat2pdf.org (http://www.pat2pdf.org) or by seeking out my AES paper on this topic.

Linear

linear

10-01-2006, 07:27 AM

i guess in combination with such startling and revolutionary design breakthroughs such as using not simply wood, but knot-free, wood for a horn, a us patent was inevitable :blink:

you have to wonder who they have working at the patent office.

Guiseppe's patent may be trivial, but a least the concept works and it protects his business interests. If you want to see the "outer limits" of what can be patented, look at 6025810 in www.pat2pdf.org (http://www.pat2pdf.org). The inventor simply claims that he is "sending the signal faster than light" due to the fact that his design "pokes a small hole into another dimension". (This is NOT a joke! Check it out.)

Linear

jack_bouska

10-01-2006, 12:27 PM

Guiseppe's patent may be trivial, but a least the concept works and it protects his business interests. If you want to see the "outer limits" of what can be patented, look at 6025810 in www.pat2pdf.org (http://www.pat2pdf.org). The inventor simply claims that he is "sending the signal faster than light" due to the fact that his design "pokes a small hole into another dimension". (This is NOT a joke! Check it out.)

Linear

Giuseppe also appears to believe that horn loaded loudspeakers alter the speed of sound in air, as quoted from the patent pdf: "Conventional horn transducers offer, if compared to dome type transducers, the fundamental characteristic of projecting the sound with greater speed through the air..."

Either something was lost in the translation from Italian to English, or maybe Giuseppe is tapping into another dimension as well.
By the way, the listening tests performed by Keith Holland confirmed that in blind listening tests of various commercial horns, the material, wood, metal, fibreglass, etc was not correlateable with perceived quality. Knotty or not!

Jack Bouska

Mr. Widget

10-01-2006, 12:45 PM

Either something was lost in the translation from Italian to English, or maybe Giuseppe is tapping into another dimension as well.
:rotfl:

You got to love multi-dimensional audio... maybe there is a new patent waiting to be had for speaker wire too. ;)

Widget

JoshK

10-04-2006, 10:37 AM

Guido

10-04-2006, 03:45 PM

Giuseppe also appears to believe that horn loaded loudspeakers alter the speed of sound in air, as quoted from the patent pdf: "Conventional horn transducers offer, if compared to dome type transducers, the fundamental characteristic of projecting the sound with greater speed through the air..."

Either something was lost in the translation from Italian to English, or maybe Giuseppe is tapping into another dimension as well.

Obviously something was lost in tanslation and I think you know exactly what the original meaning was.

English is your mother language but believe it or not there are other human beings out there with another mother language.

What additional languages you speak (write) that enable you to discuss complicated physical coherences in a forum like this?

Me 1, you?

Ok, I know english humor and I know that I as a German can't understand it :rotfl: ..... I was told that in England BTW. Nevertheless I always had a good time in England ;)

I appreciate your stunning speaker work! Your a hifi-maniac

linear

10-04-2006, 04:09 PM

Excellent job! I'd love to build a similar pair of round horn/waveguides for my system. Can I ask what lathe did you buy for this project? I have been eyeing a wood lathe for my shop and I think the biggest bottleneck seems to be the swing distance over the bed unless the head turns 90º.

I am curious if Dr. Geddes has had any trouble from Zingali with his speakers since he uses a round horn (waveguide) whose mouth is large then its depth.

It has already been suggested (in fun!) that I should have posted this project on the "Usher Heritage Forum". So, at the risk of being banished to the "Wood Turning Heritage Forum", or even the "Audio Patents Heritage Forum", I will try to answer your questions.

Lathes that cost less than thousands of dollars seem to have a max swing over the bed of 14". Obviously, this is inadequate for turning a 15.5" horn. What I did was buy the cheapest lathe available from HarborFreight Tools (http://www.harborfreight.com/cpi/ctaf/Displayitem.taf?itemnumber=45276) for $135 and modify it. In this case, the "cheapness" is an advantage, because the bed is just two square steel tubes bolted to the frame. (Quality lathes have a one-piece cast body and bed.) Thefefore, it is relatively easy to drop the bed by using 1'' "spacers". This gives a 16" swing. For the spacers, I used rectangular aluminum tube (part # 88935k551 from www.mcmaster.com (http://www.mcmaster.com)). Yes, this lathe is a piece of junk, but it did the job!

Regarding the Zingali patent, I will make some comments, but note that I am NOT a patent attorney. However, I have consulted in this field as a technical advisor and expert witness, so I know more than a little about the system. In my opinion, the Zingali patent is very weak (big surprise!), as there is almost certainly prior art published that discloses a round horn whose mouth diameter is larger than it's depth. However, since this patent has been granted, it is assumed to be valid, so the onus would be on defendent, in an action brought by Zingali, to prove in Federal court that this prior art exists and is relevant. Does Dr. Geddes even know about Zingali's patent? Perhaps not. Does Zingali know about the details about the Summa speaker? Probably not. If did did know, he would first send a "cease and desist" letter, but might not follow up with legal action. It might be too risky financially to try to enforce a "weak" patent in the courts.

In any case, what is relevant to JBL Heritage Forum Members is that it is generally accepted that one may make use the teachings of a patent for private research and construction, so long as there is no commercial use or financial gain involved. So, buy a cheap lathe, turn a horn, and (preferably) mount a JBL driver on it!

Linear

jack_bouska

10-04-2006, 08:01 PM

jack_bouska

10-04-2006, 08:41 PM

Obviously something was lost in tanslation and I think you know exactly what the original meaning was.

No, I’m not ashamed to say that I truly don’t know what the original meaning was. (and not for lack of trying to understand)

I read that paragraph in the patent several times, and could not understand what Zingali was talking about. As far as I know, horn loaded transducers do not alter the speed of sound in air. The speed of sound is a constant (depending slightly on temperature), and even the net displacement of molecules within the compression chamber and phase plug will not alter the apparent speed of sound, but rather just cause a slight Doppler frequency shift (undetectable unless at max volume).

So what do you think Zingali meant by his statement: "Conventional horn transducers offer, if compared to dome type transducers, the fundamental characteristic of projecting the sound with greater speed through the air..."?

Even in the original Italian, I believe he has just written nonsense (but feel free to correct me).

While Zingali may have built a good device, the written evidence he has included in the patent is not physically meaningful support for his claim. (eg knot free wood, depth to width profile, etc). Zingali may be a good inventor, but based on the translated patent, he is probably not a good physicist.

Zingali did not post on this forum, so it’s somewhat unfair of me to poke fun at his patent, but on the other hand, he did expose himself to some level of ridicule by making such physically indefensible claims.

As far as other languages go, as you can tell, I’m still struggling with English!

Seriously, I have been studying Spanish, but I would not be so brash as to post on a Spanish language HiFi forum unless I felt I was proficient enough to be understood, and that I had information valuable enough to share.

Guido, I understand your posts quite clearly, and I value what you write, so you’re contributions are clearly very good examples of ESL involvement in this forum.

Jack Bouska

Mr. Widget

10-04-2006, 09:07 PM

So what do you think Zingali meant by his statement: "Conventional horn transducers offer, if compared to dome type transducers, the fundamental characteristic of projecting the sound with greater speed through the air..."?I expect they are suggesting that horn drivers offer better transient response... they are just being colorful.

I doubt they are suggesting that they have developed a "sonic accelerator". :rotfl:

Widget

Guido

10-05-2006, 03:20 AM

Guido, I understand your posts quite clearly, and I value what you write, so you’re contributions are clearly very good examples of ESL involvement in this forum.

Jack Bouska

OK Jack 1:1 ;) What is ESL?

I think the patent refers to the changes in air pressure at the border from compression chamber to horn.

ralphs99

10-05-2006, 05:32 AM

Great thread!

The 'faster than light antenna' is hilarious!

Did you also see the note in the patent that 'accelerated plant growth' has been observed by the inventor and witnesses?!!
The reason, quite clearly, is that 'energy from another dimension influences plant growth'!

Awesome stuff! :applaud:

I you find any more gems like this one, please pass them along!

Cheers, Ralph

linear

10-05-2006, 07:51 AM

Zilch

10-05-2006, 12:34 PM

ESL = English (as a) Second Language....

[Short form of "EAASL" ;) ]

linear

10-08-2006, 09:55 AM

I thought that it would be interesting to compare the contour of Jack's Oblate Spheroid / Tractrtrix Waveguide with a simple Exponential Horn, like the one shown in this thread.

Assumptions:

Jack's Waveguide has a 49mm throat dia, a 200mm mouth dia, and a depth of 95mm. Let's create an Exponential Horn with the same dimensions. From simple horn theory, this will result in a flare factor, "m", of 29.6, which in turn defines a cutoff frequency of 811 Hz. This exponential horn contour is overlaid as red dots on Jack's waveguide drawing, below.

Comments:

For shallow horns/waveguides (i.e. the mouth diameter and depth are of similar size), the difference between contours is not that great. (Note that with longer horns, the difference is very significant.) In fact, if the wood turner is only a few millimeters off in his axially symmetrical oblate spheroid creation, he could inadvertantly end up with a lowly exponential horn! (or vice versa!)

Jack, on your original drawing your mention a cutoff frequency of 531.578 Hz. Is this your calculated cutoff for your waveguide, or is it a way of identifying the Tractricx section? (i.e. Is it the cutoff of a fully realized/extended Tractrix, but not this actual waveguide?)

If the cutoff frequency of Jack's waveguide is 532 Hz, versus 811 hz for an exponential horn version, then those few millimeters here and there are certainly critical!

Comments?

Linear

linear

10-10-2006, 07:11 AM

Earlier in this thread, there was considerable discussion regarding horn patents. Fun was poked at the issue and sarcasm employed. In that same spirit, it's now my turn!

[Fun/Fantasy Mode ON]

I have a friend who works for the NSA (National Security Agency), near Washington, DC, where they monitor thousands of telephone calls a day, all in the interest of national security. They also have incredibly powerful search software, that can locate "key words" (normally of a "terrorist nature") in these conversations. I asked him to search using two names (we will call them "Dr. G" and "Signor Z") and the key words "patent", "cease", "desist", "horn", and "waveguide". He immediately got a "hit". A transcript of this conversation follows.

Z: Ciao! Hello! This is _____ calling. Please can I speak to Dr. _____

G: Hello, this is Dr. ______. What's up?

Z: I regret to tell you that the horns that you are using in your speakers infringe my patent! You must cease and desist immediately! Otherwise, I will be forced to take legal action!

G: Now hold your horses there, Guiseppe! We DON'T use horns in our speakers! We use waveguides.....

Z: But waveguides are the same as horns, no?

G: Absolutely NOT! They are completely different! They are as different as chalk and cheese ... or.....or.... pasta and spaghetti!

Z: Well, maybe....let me think.....

G: Furthermore, what we use are "axially symmetric oblate spheroid CD waveguides"! This is leading edge technology, based of course, on my research! Now your exponential horns - that's old technology! They pre-date the internet - they might even be older than that!

Z: OK, OK! I apologize for my false accusation. Please accept my congratulations on your research. Perhaps I should think about upgrading my old exponential horns to the new, how do you say it, oblate sphero.....

G: You will have to license that technology from me! But you won't regret it! Nobody wants plain old exponential horns anymore. If you upgrade to my technology, your sales will skyrocket. There's room for both of us in the marketplace. The deal will be win win!

Z: OK, d'accordo! Send me a contract! But, one problem maybe? I have my computer controlled wood turning machine programmed for the exponential, you know? The cutting heads have a limited travel. For your new completely different contour, how do you say it, oblate sphero.... I may have to buy a new machine, no?

G: Don't worry! You will only have to adjust the contour a few millimeters here and there. You can use your old machine for sure! Once you sign the contract, I will give you all the details. Wait a minute, what's the tolerance on your machine?

Z: I'm not sure. It's an older machine. Probably a millimeter, here or there....

G: You are probably already infringing my intelectual property with that tolerance! I'll add a fee to the contract to cover your past production of horns. They might have oblate contours in them!

Z: OK, thank you, grazie tante! It will be great to get this new technology!

G: It's a pleasure doing business with you, Guiseppe!

[Fun/Fantasy Mode OFF]

Linear

jack_bouska

10-11-2006, 12:43 AM

I thought that it would be interesting to compare the contour of Jack's Oblate Spheroid / Tractrtrix Waveguide with a simple Exponential Horn, like the one shown in this thread................
Comments?
Linear

Hi Linear – it seems I’m inordinately busy at work these days, a condition which is seriously affecting my spare time available for posting – I will say that I have been doing some offline thinking about your horn-waveguide contour comparison post, and I am in the process of composing a reasoned reply, but my spare time “crunch” has diverted my time at the keyboard to issues more seismic oriented (same problems as audio/acoustics, just not so many octaves, and a much more complicated propagation media). – In the mean time, I am including a .jpg scan of a short sidebar article written by Dr. Keith Holland, which discusses the phenomena of horn cut-off, in relation to horn contour shape (conical vs. exponential).

I class this material as *essential* reading for anyone who may be contemplating design of their own DIY horn contour. The post I am (mentally) composing will draw on some of the information in Dr. Holland’s article, so I am posting these .jpg’s now, to give any interested parties a couple of days to read and digest the information, before I post my reply related to Geddes-Zingali differences and your particular implementation. The article may require reading in several passes, depending on your particular background level in acoustics. (I re-read it about once every year or so, just to refresh my own memory.)

Also, if anyone has not done so yet, all interested parties should download a pdf of the Peavey white paper written by C. Hughes (Google: Charlie Hughes peavey )
The Quadratic-Throat Waveguide:

http://www.installaa.com/downloads/pdf/qwp1.pdf#search=%22charlie%20hughes%20peavy%22 (http://www.installaa.com/downloads/pdf/qwp1.pdf#search=%22charlie%20hughes%20peavy%22)

Again, the Peavey paper is classed as essential reading for any aspiring horn diy’er.

All I have time for right now, more to follow, maybe this weekend, (if I first sacrifice a goat to the spare time god’s)

Jack Bouska
(thinking about a problem/question is fun, writing about it is more of a chore)

Ian Mackenzie

10-11-2006, 02:43 AM

http://home.carolina.rr.com/charliehughes/Articles/QTWaveguide/QTWaveguide-Fr.html

linear

10-18-2006, 09:21 AM

Also, if anyone has not done so yet, all interested parties should download a pdf of the Peavey white paper written by C. Hughes (Google: Charlie Hughes peavey )
The Quadratic-Throat Waveguide:

Again, the Peavey paper is classed as essential reading for any aspiring horn diy’er.

I have started working on a chart (graph) that displays several types of horn / waveguide contours, so that the differences in physical dimensions can easily be observed, should anyone want to pick one for a wood turning DIY project.

So far, I have included Conical, Exponential, and Hughes, as shown in the following attachment. The "Hughes" is from a paper by Charles Hughes (with Peavey at the time), which is referenced earlier in this thread.

Speaking of the Hughes paper, considering that it is "essential reading", has anyone else noticed that his equations have several serious typos that make them useless for working out the data points for his waveguide???? I had to go back to Pythagoras and re-derive them. I would appreciate a "sanity check" from someone else, that I'm correct in that observation.

I intend to add some more contours to this plot. I am "normalizing" everything to a horn / waveguide with a 2 inch throat diameter, a 15.5 inch mouth diameter, and a depth of 6 inches. This is my favourite size, as it is a drop in "replacement" for a standard 15 inch driver. Someone could retro-fit a cabinet with two 15" speakers, to a horn / bass driver configuration with this size of horn / waveguide.

Jack, when you get some time, could you send me an email, or post some data points (x,y) on your "Oblate Spheroid / Tractrix" waveguide? Would anyone else like to contribute their favourite contour? About 20 data point pairs should be enough, as long as the values at any "contour transistion" (e.g, Spheroid to Tractrix) are included.

Linear

Mr. Widget

10-18-2006, 10:34 AM

jack_bouska

10-18-2006, 02:58 PM

I have started working on a chart (graph) that displays several types of horn / waveguide contours, so that the differences in physical dimensions can easily be observed, should anyone want to pick one for a wood turning DIY project.

So far, I have included Conical, Exponential, and Hughes, as shown in the following attachment. The "Hughes" is from a paper by Charles Hughes (with Peavey at the time), which is referenced earlier in this thread.

Speaking of the Hughes paper, considering that it is "essential reading", has anyone else noticed that his equations have several serious typos that make them useless for working out the data points for his waveguide???? I had to go back to Pythagoras and re-derive them. I would appreciate a "sanity check" from someone else, that I'm correct in that observation.

I intend to add some more contours to this plot. I am "normalizing" everything to a horn / waveguide with a 2 inch throat diameter, a 15.5 inch mouth diameter, and a depth of 6 inches. This is my favourite size, as it is a drop in "replacement" for a standard 15 inch driver. Someone could retro-fit a cabinet with two 15" speakers, to a horn / bass driver configuration with this size of horn / waveguide.

Jack, when you get some time, could you send me an email, or post some data points (x,y) on your "Oblate Spheroid / Tractrix" waveguide? Would anyone else like to contribute their favourite contour? About 20 data point pairs should be enough, as long as the values at any "contour transistion" (e.g, Spheroid to Tractrix) are included.

Linear

Linear: so much to post, so little time

- sorry I did not get any spare typing time last weekend, or this week, so my promised horn shape reply is still in the works (I have been composing it in my head, just no time to type).

-I remember the Hughes paper had some incorrect assumptions in it, which I was planning on commenting on, but I never tried graphing a contour based on his equations, I’ll take a look over the next few days (or weeks) and get back to you.

- I’ll send you an excel spreadsheet with both my horn contour x,y’s, as well as a small program to generate oblate spheroid horn curves. Send my your normal email address by PM and I’ll package up a zip and send tomorrow.

- The Oblate/Tractrix shape needs to be built “by hand”, essentially designing a oblate spheroid horn which is shorter, and has a smaller mouth than the final horn length and diameter, then running the tractrix curve generator multiple times until the exit angle and mouth size of the oblate exactly match the throat diameter and input angle of the tractrix.

- Later this month, I will create a new hybrid (Oblate-Tractrix) contour to match your 15.5” standard, it only takes about a half hour of fiddling, but not tonight (I have a couple of weeks holiday coming, so bear with me)

- In the mean time, this (and the next four) posts will contain some graphs from a horn directivity study that I did (using the McBean hornresp program) prior to constructing the ones I am using now. With some compromise, the input Mcbean Hornresp program can be adjusted to simulate using a compression driver on a front loaded horn, which opens up the possibility of doing directivity studies using a computer, instead of a lathe.

- Eventually, I will re-run this study (with the four horn types) conforming to your 15.5”mouth, 2”throat standard, but for now the four horn types use a 1”throat, 6.3”mouth.

- The various graphs have reasonable labels for title and axes, and should be self explanatory, if you need more information, then you will have to download Mcbean hornresp, and read the help files.

- The obvious conclusion from this directivity study is that only the Oblate and the Conical shapes provide reasonable constant directivity coverage. The Exponential and Tractrix curves both show strong narrowing of directivity with increasing frequency.

More later
Jack Bouska

jack_bouska

10-18-2006, 03:01 PM

Mcbean hornresp
Horn Directivity Study: Exponential horn, frequency response and directivity curves:

jack_bouska

10-18-2006, 03:04 PM

Mcbean hornresp
Horn Directivity Study: Tractrix horn,
Frequency response and directivity curves:

jack_bouska

10-18-2006, 03:06 PM

Mcbean hornresp
Horn Directivity Study: Oblate Spheroid Waveguide,
Frequency response and directivity curves:

jack_bouska

10-18-2006, 03:09 PM

Mcbean hornresp
Horn Directivity Study: Conical Horn,
Frequency response and directivity curves:

jack_bouska

10-18-2006, 03:28 PM

This is an interesting comparison. Thanks for doing the leg work and sharing this with us. I would love for someone to produce a prototype of each horn and objectively and subjectively compare them along with a true tractrix horn and possibly one of Jack's horns.

Are Jack's "Oblate Spheroid / Tractrix" derived from any real world product or are they unique to his speakers? Has anyone ever listened to them? (Besides Jack obviously... we tend to be biased about our own babies.) Jack if you posted their full history I am sorry, I must have missed it.

Widget
The directivity study (above) shows some objective (simulated) comparisons.
In my post on Oblate Spheroid horns I show measurements which compare a tractrix horn against the pair of Oblate Spheroid designs that I built. I could comment on the sound from each, but you know that only two types of listening are valid (Either double blind scientifically controlled testing, -or- personal audition)

“Are Jack's "Oblate Spheroid / Tractrix" derived from any real world product..?”
Answer is: they are my own concoction

“Has anyone ever listened to them? (Besides Jack obviously….”

Apart from those in my immediate family, (and a good part of the neighborhood when I crank up the volume with the window’s open…..)

I’ve had four people over from the LLDIYHiFi club for a listen. A quote from the last fellow (a devout ribbon fan) was “nice treble”

Also, John W. just built a pair (see big blue thread) and so he will likely report on the sonic quality when the sawdust and solder settle on his latest project.

“Jack if you posted their full history I am sorry, I must have missed it.”
I’m sure you must have seen my thread called:
DIY Axially symmetric oblate spheroid CD waveguides, in solid Oak
At:
http://www.audioheritage.org/vbulletin/showthread.php?t=12126 (http://www.audioheritage.org/vbulletin/showthread.php?t=12126)

In that thread, I discuss (in gory detail) the waveguide contours, shape, measurements etc. In particular post #3 describes why I chose a hybrid oblate – tractrix shape for the horn mouth.

Before retiring for bed tonight, I’m going to post some additional graph comparisons on the Big Blue thread
http://www.audioheritage.org/vbulletin/showthread.php?t=12671 (http://www.audioheritage.org/vbulletin/showthread.php?t=12671)

In the big blue post, I will show comparisons of the Oblate spheroid vs the JBL PT-H95HF waveguide and a 2381. I wouldn’t go so far as to say the waveguides that I built are the be-all and end-all of sonic excellence, but they do perform, and measure just as I expected them to when they were on the drawing board (gotta like that)!

Jack Bouska

Mr. Widget

10-18-2006, 03:35 PM

“Are Jack's "Oblate Spheroid / Tractrix" derived from any real world product..?”
Answer is: they are my own concoction!

I’m sure you must have seen my thread called:
DIY Axially symmetric oblate spheroid CD waveguides, in solid Oak
At:
http://www.audioheritage.org/vbulletin/showthread.php?t=12126 (http://www.audioheritage.org/vbulletin/showthread.php?t=12126)

Thanks for the concise answer, and thanks for posting those interesting sims.

Yes, I have glanced at your thread, but I simply haven't found the time to go through the "gory detail"... based on your sims... why would you want such a contour? The on axis response is far from linear and the off axis ringing seems really problematic.

I freely admit to be in the exponential/tractrix camp. I have yet to hear a demonstration of conical, bi-radial, or other true CD horns that makes me want to go there.

Widget

jack_bouska

10-18-2006, 04:03 PM

Thanks for the concise answer, and thanks for posting those interesting sims.

Yes, I have glanced at your thread, but I simply haven't found the time to go through the "gory detail"... based on your sims... why would you want such a contour? The on axis response is far from linear and the off axis ringing seems really problematic.

I freely admit to be in the exponential/tractrix camp. I have yet to hear a demonstration of conical, bi-radial, or other true CD horns that makes me want to go there.

Widget

Widget – please re-read the post above (#43), I just finished editing it (again)

The Mcbean simulations are not perfect; most of the ringing is due to the compromises in my specification of the diaphragm and motor, along with the poor termination from the straight sided front chamber-to-throat, and the complete lack of fully curved round-over to alleviate the mouth-to-free space acoustic impedance discontinuity.

My actual measured curves don’t show anywhere near the same level of ripple, (on or off axis) as the simulations. Have a look at the directivity measurements in post #3 of my oblate spheroid thread, the off axis curves (+ - 40deg) track each other quite closely, and do not display excessive ripple. Post #4 shows some high resolution FFT graphs, (with spatial average) which also show the relatively low level of ripple from the Oblate devices). Also, if your interested, watch out for more info in my next post in the big blue thread, later tonight.

I have never used exponential horns; however they *should* provide the optimum diaphragm loading, right down to cutoff frequency (if that's required). I have built and used a couple of different Tractrix horns, on JBL 2426 and TAD2002 drivers, and found that they sounded very good, however I needed to aim them correctly, and sit within the high frequency “sweet spot” to get the full treble experience. They also worked better in over damped (Live-end Dead-end) rooms. (quite bad with wood floors and bare walls/ceilings.).

With the Oblate Spheroid horns, I get excellent frequency response regardless of what position I take in the room, as long as I am in front of the speakers. I can stand, or sit, and even walk across the entire width of the room, barely 3 feet in front of the speakers (or at the back of the room), and I hear *exactly* the same frequency response (and high frequency extension) as I would in my sweet spot favorite listening position.

I could never do that with my tractricx horns.

Hope that helps?
Jack Bouska

Mr. Widget

10-18-2006, 04:22 PM

I have built and used a couple of different Tractrix horns, on JBL 2426 and TAD2002 drivers, and found that they sounded very good, however I needed to aim them correctly, and sit within the high frequency “sweet spot” to get the full treble experience. They also worked better in over damped (Live-end Dead-end) rooms. (quite bad with wood floors and bare walls/ceilings.).
I agree... I listen in a treated room. Since I enjoy stereo, I listen primarily in the sweet spot. My current speakers have a slightly wider sweet spot than tractrix horns have, but you absolutely need to sit in the zone to get the full benefit.

With the Oblate Spheroid horns, I get excellent frequency response regardless of what position I take in the room,...For me off axis listening is just casual background entertainment and having an acceptable but less than perfect presentation there is more than adequate. As stated above, for stereo listening you have to be in the sweet spot anyway so optimum performance outside of that envelope is just not worth compromising the performance in the sweet spot.

Widget

Robh3606

10-18-2006, 04:26 PM

I can stand, or sit, and even walk across the entire width of the room, barely 3 feet in front of the speakers (or at the back of the room), and I hear *exactly* the same frequency response (and high frequency extension) as I would in my sweet spot favorite listening position.

I could never do that with my tractricx horns.

Hmmm:hmm:

I had a similar impression when listening to them as well. I like to have good balance through out the room as well.

Rob:)

jack_bouska

10-18-2006, 04:36 PM

For me off axis listening is just casual background entertainment and having an acceptable but less than perfect presentation there is more than adequate. As stated above, for stereo listening you have to be in the sweet spot anyway so optimum performance outside of that envelope is just not worth compromising the performance in the sweet spot.

Widget

It's not that I need to listen while walking around, or that I need to satisfy a widely dispersed audience, its just that I listen to the direct arrival, the early reflections, *and* the reverberant (decaying) sound field when I sit in the sweet spot.

I still think good CD is important for quality sound reproduction. Good CD means flat, ripple free (and internal reflection free) on and off axis response. The attached graphic (from JBL literature) explains it better than 1000 words from me.

Jack Bouska

Mr. Widget

10-18-2006, 04:54 PM

It's not that I need to listen while walking around, or that I need to satisfy a widely dispersed audience, its just that I listen to the direct arrival, the early reflections, *and* the reverberant (decaying) sound field when I sit in the sweet spot.

I still think good CD is important for quality sound reproduction. Good CD means flat, ripple free (and internal reflection free) on and off axis response. The attached graphic (from JBL literature) explains it better 1000 words from me.
Oh, I understand the concept, and it really makes sense. However in my own real world experience, the systems that I have really responded to, the ones that really sounded the best with stellar imaging etc. are not examples of controlled directivity. I think that in the PA world JBL is right, but so far, in domestic Hi-Fi listening, I still need to be convinced.

Widget

Robh3606

10-18-2006, 07:25 PM

Hello Widget

You like the Revels?? They are CD like as far as there in room response in concerned. More controled directivity but very similar. The power response rolls off a bit but it is very much along the same lines. Smooth curves with no sharp of abrupt changes. They are designed using the same guide lines as the 6300 series Monitors as far as On and Off Axis response.

http://www.jblpro.com/LSR/PDF/JBL.tn_v3_2A.pdf

Rob:)

Ian Mackenzie

10-18-2006, 09:19 PM

Jack,

Correct me if I am wrong but what you are talking about is the sound quality of the horn itself . Issues of driver loading, throat impediance, the horn contour and mouth termination all seem to play a role in the subjective performance whereas there are numerous biradial CD horns that have technically good dispersion but are not subjectively that great.

I read the above links, very informative.Thankyou.

Ian

Robh3606

10-19-2006, 09:31 AM

Mr. Widget

10-19-2006, 10:29 AM

Hello Widget

You like the Revels?? They are CD like as far as there first reflection. The power response rolls off a bit but it is very much along the same lines. Smooth curves that are similar with no sharp of abrupt changes. They are designed using the same guide lines as the 6300 series Monitors as far as On and Off Axis response.
Interestingly, I do like them though I am not all that crazy about the LSRs, but I don't think it has to do with their dispersion characteristics... they simply sound a bit hard at anything above easy listening levels. I know the Revel has a similar wave guide around the tweeter, but I am not sure if it would actually be considered CD. CD, like minimum diffraction designs, time aligned designs, and other previous flavors of the month may not be all that important in and of themselves... what I mean is that I have heard loudspeakers from virtually every major design theme that have sounded great... no single concept seems to be critical, though I am sure each of the various proponents would violently disagree me... it is just that if time alignment is so critical, then how can a completely screwed up system from this perspective sound fantastic? If minimum diffraction is so....

In any event, all of the CD designs that I was thinking of were horn designs, and when thinking of horn based systems, the systems that have sounded the best have never been constant directivity systems. I have heard excellent wide dispersion systems and excellent narrow dispersion systems, but those described as CD have never been even close to the best in the other categories... that may change with the advent of newer designs and technologies.

Widget

grumpy

10-19-2006, 10:42 AM

In any event, all of the CD designs that I was thinking of were horn designs, and within that family of designs, the systems that have sounded the best have never been constant directivity systems.

You lost me there. I thought CD = constant directivity? :blink: -grumpy

Mr. Widget

10-19-2006, 11:02 AM

You lost me there. I thought CD = constant directivity? :blink: -grumpyI am using CD to mean constant directivity. Where did I lose you? I rewrote the quoted line... maybe it makes more sense now?

Widget

grumpy

10-19-2006, 11:12 AM

Thanks Widget,

Requote did the trick. Not being picky intentionally... this is a useful thread of information and
experience and I want to make sure I understand the intent. Clear enough now. -grumpy

Zilch

10-19-2006, 11:32 AM

It's not that I need to listen while walking around, or that I need to satisfy a widely dispersed audience, its just that I listen to the direct arrival, the early reflections, *and* the reverberant (decaying) sound field when I sit in the sweet spot.I'm not so keen on early reflections as I once was, particularly those coming from the floor and ceiling. The ambience is nice, but it's artificial. 90° x 50° is my preferred pattern, of late, with toe-in to reduce early reflections from the front sides.

None of this diminishes the importance of uniform power response for the reverberant field, tho. Like you, it's not my desire or objective to create an "oversize headphone" listening experience, but I don't think a symmetrical dispersion pattern is the ticket, either. I particularly enjoy the H2600/H3100 asymmetric pattern.

Widget HATES it. There's no sweet spot, which is disconcerting.

[Well, there IS, but you gotta measure to find it. Tee hee.... :D ]

Robh3606

10-19-2006, 12:40 PM

I particularly enjoy the H2600/H3100 asymmetric pattern.

Widget HATES it. There's no sweet spot, which is disconcerting.

Hello Zilch

Now that's a horse of a different color. That's a DC horn or Defined Coverage as JBL called them based on the 4660.

http://www.lansingheritage.org/html/jbl/specs/pro-speakers/1984-4660.htm

I would very much like to hear what they sound like.

Rob:)

Mr. Widget

10-19-2006, 01:44 PM

Hello Zilch

Now that's a horse of a different color. That's a DC horn or Defined Coverage as JBL called them based on the 4660.

http://www.lansingheritage.org/html/jbl/specs/pro-speakers/1984-4660.htm

I would very much like to hear what they sound like."Widget hates them" is over stating it... I wouldn't pick them out to be a horn that I used on a daily basis, I certainly can see the appeal... They absolutely give a very unique stereo sound stage. If we ever have the time, it might be interesting to try out the TAD 2002 drivers on them... these drivers are supposedly good to 27KHz... I take all of these specs with a few grains of salt. One of my issues with most of the two-way designs is the quite audible distortion up top that is only aggravated by combining them with horns that have a falling off on axis response.

Widget

Zilch

10-19-2006, 01:51 PM

If we ever have the time, it might be interesting to try out the TAD 2002 drivers on them....Well, sure, we'll have the time. When the new throats get here, we can hook a pair up!

I could send some off to Rob to hear, as well, then.

[The horns, not the TADS.... :p ]

One of my issues with most of the two-way designs is the quite audible distortion up top that is only aggravated by combining them with horns that have a falling off on axis response.I believe you will find 2452H-SL on 2352 very different in these respects.

Not bad on the PT Waveguide it was designed for, either, which is on topic, since Jack cited PT-H95HF above.... :)

Robh3606

10-19-2006, 02:35 PM

One of my issues with most of the two-way designs is the quite audible distortion up top that is only aggravated by combining them with horns that have a falling off on axis response.

Hello Widget

Sounds like you are partial to horns that have changing DI to compensate on axis and do not like using any electrical compensation. The horn does the EQ so there is no monkey business over all with the compression driver output. Just some resistive padding to level match. CD's certainly would not be your cup of tea. We all have our like and dislikes and it's what you like that matters. There is no absolute right or wrong. There are so many ways to do things with speakers in general that makes it a lot more fun than the same damn thing over and over again. Your right I have heard all different aproaches sound really good so you have to wonder what are they doing "right" in each of those unique designs and try to figure out what makes them standouts from others that share the same approach. Some just seem to have that edge where the whole is better than the sum and they really shine. It just all comes together and when you hear it you can't miss it.

Rob:)

Mr. Widget

10-19-2006, 02:45 PM

Sounds like you are partial to horns that have changing DI to compensate on axis and do not like using any electrical compensation. The horn does the EQ so there is no monkey business over all with the compression driver output.So far... I am open to being surprised though.

...I have heard all different aproaches sound really good so you have to wonder what are they doing "right" in each of those unique designs and try to figure out what makes them standouts from others that share the same approach. Some just seem to have that edge where the whole is better than the sum and they really shine. It just all comes together and when you hear it you can't miss it.Yeah... I haven't found the common denominator... it is surprising how so many different approaches can lead to outstanding reproduction... then again, the field of losers is littered with far more speakers...:D

Widget

linear

10-23-2006, 02:57 PM

I have started working on a chart (graph) that displays several types of horn / waveguide contours.............

I intend to add some more contours to this plot............

I have started a new thread, "Horn / Waveguide Contour Comparisons", on the DIY Forum, to share information about the shape and performance of the various horns and waveguides that may be of interest to DIYers. Check it out at:

http://audioheritage.org/vbulletin/showthread.php?p=130339#post130339

Hopefully, this will be more convenient and focused on topic than this thread.

Linear

linear

10-23-2006, 03:35 PM

I just re-read the Zingali horn patent and noted that the independent claim is: "......comprising single horn.......circular cross section.....and a conic exponential profile, the axial length of said diffuser being less than the maximum diameter thereof."

Therefore, only horns with an exponential contour are covered by his patent! So, Dr. Geddes and Dr. Edgar are NOT infringing Zingali with their Oblate and Tractrix contours, respectively.

Sorry for stirring up the hornets nest (horn...ets, get it?).

Linear

linear

10-31-2006, 08:19 AM

Hello Jack

I have a question about both the Peavey horn and Earls. In both cases the measurements stop at 10K. What actually going on above that frequency?? The 2344 used the difraction slot width to determine the HF beamwidth limit in the horizontal plane. Both Earls horn and the Peavey don't have this feature so the smallest dimension is the throat diameter. In Earls I think it's a 1" throat while the Peavey is 1.6". Are the throat dimensions acting as a difration slot as far as the upper limit dispersion in concerned. Does the directivity change above 10K???

Rob:)

Good question.

In his white paper on his Summa Loudspeaker (http://www.gedlee.com/downloads/Summa.pdf), Earl Geddes says:

"Gedlee does not test above 10 kHz beause there is virtually no important musical content at these extreme frequencies, preferring instead to concentrate on what is important - the mid band 100 - 10,000 Hz. There is still significant energy output above 10 kHz in this system, it's just not shown because it is not important."

This seems like a rather radical position to take. Is Dr. Geddes saying that all "super tweeters" deliver "no important musical content"? Are JBL and the Project May Team wasting their time and money on the 045Be transducer?

Maybe I'm overly suspicious, but in my experience radical statements are often used to mask inadequacies. So, does anyone really know what is going on with Earl's waveguide above 10 kHz?

Linear

Mr. Widget

10-31-2006, 10:08 AM

"Gedlee does not test above 10 kHz beause there is virtually no important musical content at these extreme frequencies, preferring instead to concentrate on what is important - the mid band 100 - 10,000 Hz. There is still significant energy output above 10 kHz in this system, it's just not shown because it is not important."
Maybe the good doctor is deaf.

I suppose he can make his argument based on the science of music, using all of his math and theories, but all it takes is a quick listen to realize that there is more going on than what we typically can discern from these techniques.

I haven't done any experiments along the lines of the DD66000 where there is a supertweeter coming in at 20KHz, but changes to your system above 10KHz can be quite dramatic.

Widget

scott fitlin

10-31-2006, 10:13 AM

Maybe the good doctor is deaf.

I suppose he can make his argument based on the science of music, using all of his math and theories, but all it takes is a quick listen to realize that there is more going on than what we typically can discern from these techniques.

I haven't done any experiments along the lines of the DD66000 where there is a supertweeter coming in at 20KHz, but changes to your system above 10KHz can be quite dramatic.

WidgetHell yeah we can hear above 10k, and it does make a difference to what you hear. I turn the 16k eq fader down, I DO hear a difference!

Even if there were no musical content up that high, these frequencies do in fact, affect things going on lower in frequency that we really can hear.

So, Im in agreement with Widget.

:D

TimG

10-31-2006, 08:47 PM

When I was experimenting with different crossover frequencies with the Project May drivers, I tried crossovers with and without the 045Be. Even though the 435Be had extension to 13kHz and beyond, and even though I shouldn't be able to hear past 16khz, the difference in sound with and without the 045 was obvious to me and anyone else who heard the system. Maybe Earl is not interested in response above 10kHz because with a 1" throat horn you are no longer going to get controlled directivity much above that point. The dispersion at that short of a wavelength, or slightly above, is already determined within the phase plug of the driver with a 1" throat so maybe he doesn't feel that it should receive attention. Also, trying to integrate a supertweeter at 10kHz, unless it was mounted concentrically within the larger compression driver, and digitally time aligned, come along with the compromise of introducing comb cancellations in the response due to the small wavelengths involved. I'm not saying it can't still sound good and add to the experience, but it still presents problems.

jack_bouska

11-04-2006, 09:55 AM

Good question.

In his white paper on his Summa Loudspeaker (http://www.gedlee.com/downloads/Summa.pdf), Earl Geddes says:

"Gedlee does not test above 10 kHz beause there is virtually no important musical content at these extreme frequencies, preferring instead to concentrate on what is important - the mid band 100 - 10,000 Hz. There is still significant energy output above 10 kHz in this system, it's just not shown because it is not important."

This seems like a rather radical position to take. Is Dr. Geddes saying that all "super tweeters" deliver "no important musical content"? Are JBL and the Project May Team wasting their time and money on the 045Be transducer?

Maybe I'm overly suspicious, but in my experience radical statements are often used to mask inadequacies. So, does anyone really know what is going on with Earl's waveguide above 10 kHz?

Linear

I have a *long* thread in preparation which discusses these, and other issues, hope to get it posted within the next few days. - Jack

Mr. Widget

11-04-2006, 12:15 PM

I have a *long* thread in preparation which discusses these, and other issues, hope to get it posted within the next few days. - JackMaybe you could post a Cliffs Notes version too?:D

Widget

Titanium Dome

11-04-2006, 01:48 PM

Narrowing the band like that has to reduce the scope of work. As part of an over all work plan, it's brilliant, but in this case seems like a little too much maverick spirit given the state of scientific and technical accomplishment in the areas of aural perception and audio reproduction.

jack_bouska

11-07-2006, 09:08 AM

I have completed my long reply to the questions from Ian Mackenzie and Rob H, and have placed these in a new thread called:
Factors Affecting Sonic Quality of Mid & HF Horns & Waveguides
http://www.audioheritage.org/vbulletin/showthread.php?t=12967 (http://www.audioheritage.org/vbulletin/showthread.php?t=12967)

Regarding Widget's request for Cliff's notes, those who might be in a hurry are invited to read the first couple of paragraphs of the opening post, then skip down to the conclusions at the end of post 9. to save time.

Otherwise, the text and graphs in between are only of interest if you want to understand more about system design, horn behavior, and perhaps learn a few techniques which can be used to make your next horn based DIY project sound a little better.

Jack Bouska

ChopsMX5

08-20-2007, 09:24 AM

Handmade, part M9500, part Zingali, and part D2.

http://www.audioheritage.org/vbulletin/attachment.php?attachmentid=18953&stc=1&d=1159167109

I hate to bring up an old thread, especailly since this is my fist post here, but has it ever been told how this particular setup sounds?

The reason I ask is because I am going to be building a very similar setup in a couple of week, but with Altec 511B's and 902-8B drivers. And since my room has limited space, building 3 identical loudspeakers (one being horizontal for the center channel) at Usher D2 specs - 10cf tuned to 28Hz, I won't have room for subwoofers. From what I've read elsewhere, they claim the D2 is good for low 20's or high teens as far as bess extension is concerned. If this is the case, then I wouldn't need a sub(s) anyway.

Here's a quick overview of my current setup, with lots of nice pictures as well! :D

http://www.diyaudio.com/forums/showthread.php?s=&threadid=91372&perpage=25&pagenumber=11 (http://www.diyaudio.com/forums/showthread.php?s=&threadid=91372&perpage=25&pagenumber=11)

Linear, if you're around, I'd sure like to hear what you have to say about your setup.

Many thanks in advance!

Charles

linear

08-20-2007, 08:13 PM

Hello Charles,

Thanks for your interest in my loudspeakers. Now is a good time to purchase 15HM drivers, as they are on sale at www.partsexpress.com (http://www.partsexpress.com) for $138.00 and free shipping. These are exactly the same speakers that Usher use in the D2.

However, the D2 is designed to be a high efficiency / low distortion speaker, and its bass extension will NOT be as good as a top of the line subwoofer. Please see the plot below.

The red curve is the 15HM in a 10cf cab tuned at 28 Hz. The other Usher curve is my smaller upper cab tuned to 35 Hz. The JBL curves are for their M9500 with their speakers (1400 Nd) and volumes. These curves are NOT measured, but caluclated with the best T-S parameter data that I could find.

Note that the Usher with the 10cf cab is the best, but not exactly in the subwoofer class. Bottom line, I think that you might miss your subwoofers on your pipe organ and home theatre "earthquake" material.

Good luck with your project.

Linear

I hate to bring up an old thread, especailly since this is my fist post here, but has it ever been told how this particular setup sounds?

The reason I ask is because I am going to be building a very similar setup in a couple of week, but with Altec 511B's and 902-8B drivers. And since my room has limited space, building 3 identical loudspeakers (one being horizontal for the center channel) at Usher D2 specs - 10cf tuned to 28Hz, I won't have room for subwoofers. From what I've read elsewhere, they claim the D2 is good for low 20's or high teens as far as bess extension is concerned. If this is the case, then I wouldn't need a sub(s) anyway.

Here's a quick overview of my current setup, with lots of nice pictures as well! :D

http://www.diyaudio.com/forums/showthread.php?s=&threadid=91372&perpage=25&pagenumber=11 (http://www.diyaudio.com/forums/showthread.php?s=&threadid=91372&perpage=25&pagenumber=11)

Linear, if you're around, I'd sure like to hear what you have to say about your setup.

Many thanks in advance!

Charles

ChopsMX5

08-20-2007, 08:58 PM

Hello Charles,

Thanks for your interest in my loudspeakers. Now is a good time to purchase 15HM drivers, as they are on sale at www.partsexpress.com (http://www.partsexpress.com) for $138.00 and free shipping. These are exactly the same speakers that Usher use in the D2.

However, the D2 is designed to be a high efficiency / low distortion speaker, and its bass extension will NOT be as good as a top of the line subwoofer. Please see the plot below.

The red curve is the 15HM in a 10cf cab tuned at 28 Hz. The other Usher curve is my smaller upper cab tuned to 35 Hz. The JBL curves are for their M9500 with their speakers (1400 Nd) and volumes. These curves are NOT measured, but caluclated with the best T-S parameter data that I could find.

Note that the Usher with the 10cf cab is the best, but not exactly in the subwoofer class. Bottom line, I think that you might miss your subwoofers on your pipe organ and home theatre "earthquake" material.

Good luck with your project.

Linear

Hi Linear,

Thanks for getting back with me! I really appreciate it.

Yep, PE is where I plan on buying them. In fact, in just a few minutes I'll be placing the order!! Although, everyone at the moment is selling the 15HM's at that price, so it must be a manufacture special.

In your plot, it's the same exact reading I got in WinISD Pro. Ovbiously, it would be since the T/S parameters and enclosure specs are the same. Still, at 1 watt input, it says 82dB @ 20Hz for ONLY one driver.

Neither one of these plots take into account room gain or the fact that there will be THREE more identical drivers and enclosures in the same room supporting eachother. Once you add the other three drivers, you're up to 90-91dB @ 20Hz, then when you add in room gain, that's about another good 4-6dB, give or take... BTW, room gain in my room comes in strong in the sub-30Hz range, which is a really good thing.

On top of that, that's still not including the two other enclosures that make up the center channel. However, I think I will have them crossed over at 50Hz instead of playing fullrange. I've found that when you have the center set to fullrange, most of the bass goes to it instead of the mains. I'd rather that bass for to the main channels.

I guess all I can do at this point is just build them and see what happens. I must say though, even with the Cornwalls playing full range in my room (they're tuned to 37Hz), the low pedal notes can still somewhat be heard and felt. So with that in mind, I think 4 cabinets tuned 10Hz lower with twice the woofer-age might just pull this off.

So, what about your setup? You still haven't mentioned how they sound, how their bass extension is in your room, etc, etc...

Thanks again,
Charles

linear

08-21-2007, 10:15 AM

Charles,

I agree with everything that you say, including "I guess all I can do at this point is just build them and see what happens."! Please let us all know how everything works out for you.

Regarding how my particular setup sounds, well that's a very subjective thing! My wife really likes them, and she would have to, to even allow these monsters in our living room. To try to get more specific, brass and woodwinds (saxophones, trumpets, clarinets, etc) sound fantastic with the wooden horns. The classic Swedish session "Jazz at the Pawnshop" never sounded better to me. You can close your eyes and swear that you are right there in the "Stampen" Club! In this regard, I can highly recommend the new XRCD and SACD pressings available from www.fimpression.com (http://www.fimpression.com) .

Now, regarding bass performance, I'm a musician (electric bass player, actually) and I really like how my system sounds with contemporary (rock, jazz, etc) "low end". For example, the "remixed" version of "Come Together" from the new Beatles "Love" CD is just amazing. (This is George Martin's favorite Beatle's track, according to the liner notes.) McCartney's bass is swampy, smooth, but atriculate......just absolutely great. I wish I could get that exact sound myself, but somehow I just can't seem to get the "slide" spot on!

However, a few years ago, I purchased a couple of the JBL 1500 SUB speakers that Parts Express were "blowing out". I installed them in cabs almost identical to the first Revel subwoffers and I drive them with almost 1000 watts each, through an electronic equalizer that I built. I use this rig for my bass guitar (Steinberger L2), but once I did try them as a subwoofer add-on to the system described in this thread. They certain DID add more low end, but mostly it was only noticeable on "earthquake" or 16 Hz organ pedal stuff. I'm not really into home theater or organ very much, so the D2 type system suits for our listening tastes. (There's no room for the subs anyway, and my wife would kill me if I were to suggest it as a permanent thing!!)

Again, please let us know how it all sounds to you, Charles, when you finish your project.

Linear

Hi Linear,

Thanks for getting back with me! I really appreciate it.

Yep, PE is where I plan on buying them. In fact, in just a few minutes I'll be placing the order!! Although, everyone at the moment is selling the 15HM's at that price, so it must be a manufacture special.

In your plot, it's the same exact reading I got in WinISD Pro. Ovbiously, it would be since the T/S parameters and enclosure specs are the same. Still, at 1 watt input, it says 82dB @ 20Hz for ONLY one driver.

Neither one of these plots take into account room gain or the fact that there will be THREE more identical drivers and enclosures in the same room supporting eachother. Once you add the other three drivers, you're up to 90-91dB @ 20Hz, then when you add in room gain, that's about another good 4-6dB, give or take... BTW, room gain in my room comes in strong in the sub-30Hz range, which is a really good thing.

On top of that, that's still not including the two other enclosures that make up the center channel. However, I think I will have them crossed over at 50Hz instead of playing fullrange. I've found that when you have the center set to fullrange, most of the bass goes to it instead of the mains. I'd rather that bass for to the main channels.

I guess all I can do at this point is just build them and see what happens. I must say though, even with the Cornwalls playing full range in my room (they're tuned to 37Hz), the low pedal notes can still somewhat be heard and felt. So with that in mind, I think 4 cabinets tuned 10Hz lower with twice the woofer-age might just pull this off.

So, what about your setup? You still haven't mentioned how they sound, how their bass extension is in your room, etc, etc...

Thanks again,
Charles

ChopsMX5

08-21-2007, 04:02 PM

Hi Linear,

I certainly will let you ALL know how this project turns out. It's funny that you mention that you play electric bass and that you'r a musician. Back in middle school and highschool, I played string bass. Four of those six years, I was 1st chair bass, the last three years of which was a rather large 101 piece orchestra. Boy, those were the good ole' days (13+ years ago). In the hopes of re-discovering my musical tallent, I purchased an Ibanez 5-string bass about a year ago. I was able to pick it up and start playing along to various CDs right away, just by ear. I thought I'd enjoy playing jazz or blues, but apparently, it's not my thing. I'd much rather be playing classical music with a string bass in a large orchestra again. So needless to say, the bass just sits there now. :(

Anyway, I wonder how much low end you're missing due to only have two 10cf enclsoures tuned to 27Hz, and the fact that they aren't in or near the corners of the room.

See, not only will I have four 10cf enclosures, all tuned to 27Hz, but I have no choice but to place them smack-dab in the corners of the room. So who knows, I might just get the bass extension I'm looking for.

I'm also going by what Bill from Response Audio said about the Usher D2 system when he had them in his showroom. He said that he had measured their response down to the lower 20's/upper teens, and that was powering them with flea-powered SET amps. Since the drivers I'm getting are the same and the enclosures will roughly be the same (although he says the D2's are 7cf each, not 10cf), and he had them pretty much positioned similar to your setup (away from the corners), there's a very good possability of getting some extra deep bass out of these things.

If anything, I can give them a little boost with my EQ if need be (although I'd rather not), and if I decide to use my Crown XTi 1000 amp, that's 500 watts per channel into 4 ohms. That would be more than plenty oomph to motivate them.

We'll just have to wait and see what happens. Obviously, I'm very anxious to get the ball rolling on this project, and even more anxious to get it finished and running.

BTW, I just wanted to ask a couple more questions about your speakers...

What does the staggered tuning do for the sound in your system rather than going the route I am, besides the fact that you were going after the M9500 design?

What kind of accoustic treatment did you use inside your enclosures, foam, stuffing, both? And how about bracing?

Again, I'll keep everyone here updated on this project as well as document the entire build. In fact, last night I put in the oder for the 4 Usher drivers, 6 pairs of satin nickle plated binding posts (the ones that go straight through the wood), and finally, that Jazper circle cutting adapter for the router.

Thanks again Linear for shedding some light on this subject. I really appreciate all the information you have provided!

Charles

A9X

08-22-2007, 11:30 AM

I use this rig for my bass guitar (Steinberger L2)Another bassist. My main axe is an XL2.
Great speakers.

linear

08-23-2007, 01:08 PM

A9X,

My L2 was one of the last made before they switched to the XL2. If you're not already a member, you should join the Steinberger yahoo group:

http://launch.groups.yahoo.com/group/steinberger_world/

Great info and discussion.

Now back to topic.

Charles,

Yes, you are right. I am loosing some low end response due to the fact that the upper cabs are smaller. Originally, I wanted to emulate the JBL M9500 system. JBL marketing claimed (at the time) that stagger tuning, or "imaginary Equivalent Tuning" (IET), as they called it, provided a "better balance between the high speed of Bessel tuning and the flat frequency response qualities of Butterworth tuning".

Also, at that time, Greg Timbers, the JBL design engineer involved with this IET project, "felt that by having different volumes and different tunings, we could spread the various enclosure and tuning resonance frequencies over a (wider) range, making them less of a problem".

However, there is the school of thought that all these resonances just merge into a single 4th order system anyway. Mr. Timbers was kind enough to post his "latest" thoughts, about all this, on this forum about 2 years ago. Please see:

http://audioheritage.org/vbulletin/showthread.php?t=7564&highlight=iet

Regarding construction details on my speakers, the smaller upper cabs have one brace, running side-to-side at about the middle. The larger bottom cabs have a full "shelf", with two large holes in it, at the point where it's volume equals the top cab. The volume above this "shelf" is around the wooden horn and driver, and brings the volume up to about 10 cf. There is also a side-to-side brace across the bottom half of the bottom cab. Everything is constructed out of 1 inch thick MDF.

Originally, I lined all internal walls, except the front face, with about 2" of pink fiberglass. However, this ADDED to much additional effective volume to the cabs, and changed the tuning. So, I cut the fiberglass back to about 1" and everything measured OK at 28 Hz and 35 Hz for the large and small cabs respectively.

Linear

ChopsMX5

08-24-2007, 06:03 PM

Charles,

Yes, you are right. I am loosing some low end response due to the fact that the upper cabs are smaller. Originally, I wanted to emulate the JBL M9500 system. JBL marketing claimed (at the time) that stagger tuning, or "imaginary Equivalent Tuning" (IET), as they called it, provided a "better balance between the high speed of Bessel tuning and the flat frequency response qualities of Butterworth tuning".

Also, at that time, Greg Timbers, the JBL design engineer involved with this IET project, "felt that by having different volumes and different tunings, we could spread the various enclosure and tuning resonance frequencies over a (wider) range, making them less of a problem".

However, there is the school of thought that all these resonances just merge into a single 4th order system anyway. Mr. Timbers was kind enough to post his "latest" thoughts, about all this, on this forum about 2 years ago. Please see:

http://audioheritage.org/vbulletin/showthread.php?t=7564&highlight=iet

Regarding construction details on my speakers, the smaller upper cabs have one brace, running side-to-side at about the middle. The larger bottom cabs have a full "shelf", with two large holes in it, at the point where it's volume equals the top cab. The volume above this "shelf" is around the wooden horn and driver, and brings the volume up to about 10 cf. There is also a side-to-side brace across the bottom half of the bottom cab. Everything is constructed out of 1 inch thick MDF.

Originally, I lined all internal walls, except the front face, with about 2" of pink fiberglass. However, this ADDED to much additional effective volume to the cabs, and changed the tuning. So, I cut the fiberglass back to about 1" and everything measured OK at 28 Hz and 35 Hz for the large and small cabs respectively.

Linear

Hi Linear,

I can see where one would think of the staggered design helping in that way. And thanks fot the link to the other thread. That was a good read as well!

So everything is 1" MDF, even the baffles. I assume the enclosures are plenty ridged enough then. Since these drivers have such a low Xmax, I guess they don't put too much stress on the enclosures.

When I build mine, I plan on using joining biscuits in all the joints along with small strips of wood running the entire length of each joint. Then I'll have 2 verticle shelf braces behind the drivers. The horn enclosures will be built in a similar fashion, minus the shelf braces, so it can support the weight of the top bass cabinet.

Do you think it is absolutely necessary for the front baffles to have edge bevels, especially considering the fact that the drivers will be surface mounted? I guess that might be a stupid question since both Usher and JBL have edge bevels on the D2's and M9500's. If you completely ignore this question, I'll know why. ;)

Thanks,
Charles

BTW, the Usher 15HM drivers came in today. I took one out of the box just to check it out... WOW, these suckers are massive! Very nice, well made drivers. Sure, they may not be JBL's, but they aren't cheap junk either. I just hope I can get them to perform as good as they look!

linear

08-25-2007, 08:10 PM

The edge bevels are probably more of a cosmetic thing than anything that will dramatically affect acoustic performance.

The easy way to do the bevel is to use two 1/2" thick pieces of MDF for the front baffles, cut a hole in each, of the appropriate different diameter, and them glue and clamp them together. Voilà, a 1" thick front baffle with a bevel.

The 15HMs are good speakers for the money. However, if anyone on this board, or at JBL, want to give me four JBL 1500AL speakers (like was done with Project May), I will swap out those 15HMs in a heartbeat!!!!!:bouncy:

ChopsMX5

08-25-2007, 08:22 PM

I was kind of thinking that the bevels were a little more cosmetic than anything else, but I didn't want to say it. Doing without it will definately ease the build.

I'm just banging my head on what to do as far as design goes. Over on the DIY Audio site, there's a pretty knowledgable guy there by the name of GM. He's been shooting back and forth with me about this project, and is telling me that an MTM is not the best choice for my 511B horns and that I would be better off with the 15's horizontally opposed like the 4435 or K2 Everest. Something about the 511's inability to match the 15HM's polar response in an MTM setup.

So it seems that the final design is still in the air until I can get something figured out. Hmm...

Zilch

08-26-2007, 12:22 AM

http://www.altecpro.com/pdfs/vintage/SpeakerAndMics/horns/811B%20HF%20Horn.pdf

ChopsMX5

08-26-2007, 05:50 AM

http://www.altecpro.com/pdfs/vintage/SpeakerAndMics/horns/811B%20HF%20Horn.pdf

Hi Zilch,

I have these data sheets downloaded from over a year ago when I was thinking about buying the Altec horns and drivers. Unfortunately, I don't understand some of those charts.

If you could explain to me what the charts say and how the polar response of the 511's would match up to the 15HM's, I would really appreciate it a lot.

Thanks,
Charles

Zilch

08-26-2007, 11:14 AM

Ian Mackenzie

08-26-2007, 02:29 PM

[quote=Zilch;182909
It's clearly not just a matter of throwing two woofers in a box and hooking them up.[/quote]

Who said it ever was....:D...where's Giskard when you need him....:yes:

ChopsMX5

08-26-2007, 03:20 PM

Go to the Eargle sound reinfocement manual on the JBL Pro website, in the technical library, and look up the chart on beamwidth of woofers. It varies with frequency and driver diameter. What's the beamwidth of a 15" woofer at your crossover frequency?

Then look at the 511 beamwidth plot. What's the horizontal beamwidth at that same frequency, and are they comparable?

I look at the 511 plot, and at 1 kHz, it's in the range of 90° - 100°, and JBL's 4430 used 15" 2235H at that frequency, stating that 100° was a good match.

You need to do a similar comparison for your design.

From the Project May design work, it's apparent that MTM is not a slam dunk, and from my own experience, neither is dual woofers horizontal. Look at the bandwidth the "Helper" woofer in E2 is playing, and how it mates with the main to accomplish a bass extension design objective. It's clearly not just a matter of throwing two woofers in a box and hooking them up.

I looked around on the Pro site for a quick few, but didn't find the chart you were talking about. I'll go back and look some more, unless you have the direct link. Does it mention the chart(s) anywhere in the titles?

And what's this "helper" woofer you're referring to in the E2? In fact, what is the E2?

And yes, I know there's a lot more involved that just throwing a couple of drivers in a box. I don't come off looking that hopeless, do I? :blink: