Have legacy audio got it right?

[IMGDEAD]http://www.legacy-audio.com/2004/images/helixp.jpg[/IMGDEAD]

For my next upgrade I've been thinking of ways to get more output, lower distortion, and better room interaction.

I've been interested in larger HE coaxials like the Adire HE10/12 but the treble just isn't sweet souding like what I have now.

What appeals about the large coax approach is the directivity control, and I have Ultracurve to get them flat.

But with a digital xo like Ultradrive, perhaps a sweeter sounding result could be achieved with the advantages of a large coax, using an "array" like that used in some Legacy Audio speakers. A central tweeter with a waveguide could achieve 95 db sensitivity with a number of mids around it.

How does this compare to a line array?

What is your take on how to approach this with the following goals:

* high output (pref high efficiency and power handling)
* directivity control
* sweet sounding
 
high output and directivity... sounds like a horn

Sweet sounding, well some may or may not agree with horns having that quality.


I think 3 mids around a tweeter might be better than 4(at least it would look better IMO). Placemet could be in a triangle with the tweeter in the small middle space
 
How does this compare to a line array?

Greets!

Apples n' oranges. A focussed circular array is about creating a super large duplex, ie. point source. Many moons ago I 'ringed' a 500 Hz 90 x 40 horn with four HE 12s, first on a flat baffle, then on a bifurcated one to better match (focus) the 12s to the horn's response at the XO point. Large, expensive, and pretty incredible. Had I been more inventive, I would have built a large conical horn for it, BP'd the 12s and beat TD to the Unity patent by a few decades. ;)

Scaling this down a bit, a WG loaded tweeter and some 8s could be nice, but what 'sweet' sounding dome tweeter has 95 dB sens.?

GM
 

AJinFLA

Banned
2005-02-09 4:35 am
Tampa
This one, if you don't mind the 3-4ohm impedance. I'm going to use it in a WG for a point source quite different from the one you are referring to.
http://www.d-s-t.com/link/scs/data/d2904_710000d.htm
It might be the best dome currently available. Some pics of the current WGs in the stable. Only one is a point source at this time.


Cheers,

AJ
[IMGDEAD]http://www.openbafflespeakers.com/images/Orion_Fr_CloseupWeb3.JPG[/IMGDEAD] [IMGDEAD]http://www.openbafflespeakers.com/images/PSDipole_web011.jpg[/IMGDEAD]
 
high output and directivity... sounds like a horn

Yes, it does, but it's not the only way to achieve this. Some prefer 15" drivers for their directivity and pattern control for midrange, which matches that of a horn tweeter more readily.

Thanks for the link Rudolf, interesting. Looks like a good design, although I think they need more baffle depth for the sealed mids to kill the rear wave.

bifurcated:

"Forked or divided into two parts or branches, as the Y-shaped styles of certain flowers."

(and I thought it sounded like something a cat would do)

GM, why add on a tweeter if it was the HE12?

What did you do with it in the end?

GM, I can't believe you don't have a website, having done so much! (and I've seen you appear on just about every forum there is)

AJ, my hip pocked flinches at the sight of the SS tweeter! Looks like a non phaseplug version of the ring radiator tweeter! Looks very impressive though. What are your thoughts on it?

Very interesting looking project. I'm puzzled why put a waveguide on a 95db sensitive tweeter to go with a conventional mid driver! Active speaker? Tell us more or is there a thread about it somewhere?

Is the the Seas P17rex coax?
 

AJinFLA

Banned
2005-02-09 4:35 am
Tampa
Hi Paul,

I believe GM would need a tweeter because "HE" 12 means "high efficiency" 12" woofer. Still need a tweeter. A high efficiency one to closer match the (very) high efficiency created by 4! woofers. The WG would not increase the efficiency of the tweeter, except in the lower ranges, where it loads the driver by several db, depending on the shape, dia, etc of the guide. What it would do is create controlled directivity of the driver, which I believe is crucial for the duplex to work. Thats why Donald uses one in his design.
BTW by Donald I mean this person (I believe the HTGuide thread Rudolf posted was started by someone who saw this thread) http://www.madisound.com/cgi-bin/discuss.cgi?noframes;read=348123
Unfortunately Donald does not go into any details about his design (its being patented). Perhaps GM would be kind enough to explain how this works. I would have to imagine that it approximates a point source, since its still multiple drivers mounted in different points in space.
Ok Paul, back to the SS. Yes! It makes my hip pocket flinch also! I have not dropped the $700USD on them yet ( I can't do such things on a whim), but, according to Ken Kantor (who posts on Mad as well and was kind enough to correspond with me about WGs), their (Tymphany/DST) engineers pulled all the stops to create the D2904. The 710000 is the non phase plug version, which appears to have superior dispersion (something that is a must for me) all the way up. I dont have 1st hand experience or any data other than theirs, but I still beleive (call it a hunch) that this is possibly the finest tweeter available. The $700 will come soon enough;).
I think you may have my projects mixed up. The picture I posted is a Vifa XT-19 loaded by a wavegude atop a W22. No 95db efficiency there! Again a WG is not a conventional horn. It has no effect on the efficiency of the driver except in the lowest freq's. Yes it is active. It an Orion clone with my own twist.
Thats not a Seas (although good guess, since Seas licensed the tech from KEF), it a KEF Ci200QR. Much closer to a real point source (although not a true one) than a duplexed array IMHO. Again hopefully GM can explain, since I'm not clear on how it all works and could be dead wrong.
Lastly (work beckon's), here is the awaiting recipient of the SS/WG. A Dynaudio MW190. You can guess where its going to go:) . My idea of a "point source". Nothing like the Legacy or the DNA.

Cheers,

AJ[IMGHTTPDEAD]http://www.openbafflespeakers.com/images/pyle_004.jpg[/IMGHTTPDEAD]
 
Greets!


> This one, if you don't mind the 3-4ohm impedance.

>It makes my hip pocket flinch also! I have not dropped the $700USD on them yet
====
Thanks! At this price/BW though, I would rather use an alum. diaphragm loaded 1" exit compression driver/CD lens. A whole lot more performance and probably costs less to boot.
====
>bifurcated:

"Forked or divided into two parts or branches, as the Y-shaped styles of certain flowers."
====
Right, the top two 12s were angled down and the bottom two angled up. What I forgot to mention is that when using a round or square tweeter WG, the baffle would be quadfurcated to mimic a circular dispersion.
====
>GM, why add on a tweeter if it was the HE12?
====
AJ pegged it. I did this back in the '60s with Altec VOTT components, so the equivalent of Adire's HE 12 was Altec's 601 duplex.
====
>What did you do with it in the end?
====
Nothing. I'm a curious sort, so I cobbled together countless dozens of 'proof of concepts' to satisfy my curiosity/educate myself. The answers probably were/are in textbooks, but higher math is as indecipherable to me as Sanscrit, so I had to learn most things the hard way.
====
>GM, I can't believe you don't have a website, having done so much! (and I've seen you appear on just about every forum there is)
====
Actually, I've had three, but for various reasons beyond my control they disappeared, so after 'strike three' I said to heck with it.
====
>A high efficiency one to closer match the (very) high efficiency created by 4! woofers.
====
Yep, just running the numbers to correct to 1 m/2.83V, the 'speaker' was ~105 dB in its ~flat passband.
====
>The WG would not increase the efficiency of the tweeter, except in the lower ranges, where it loads the driver by several db, depending on the shape, dia, etc of the guide.
====
Yep.
====
>What it would do is create controlled directivity of the driver, which I believe is crucial for the duplex to work.
====
Absolutely! The problem with typical duplexes though is that the horn doesn't go low enough to blend well without upsetting the midbass response, even with a 15" frame, so finally Altec 'bit the bullet' and made an optimized duplex, but due to a rather long, sad story it unfortunately was mostly a stillborn project: http://www.lansingheritage.org/html/altec/specs/pro-speakers/9864.htm I had planned to possibly resurrect it as a DIY kit of sorts, but then along came the Unity, rendering it just another footnote in audio history so didn't see much point in it.

A new company, Iconic, is marketing a tweaked version of the last Altec 604 duplex and it's been mentioned on the Altec forum that if they are financially successful enough they will eventually come out with a larger HF horn version of their 704-8A, though I imagine it will probably be limited to a 1 kHz XO point like the original JBL/Altec design.
====
>Unfortunately Donald does not go into any details about his design (its being patented). Perhaps GM would be kind enough to explain how this works. I would have to imagine that it approximates a point source, since its still multiple drivers mounted in different points in space.
====
I didn't read all the responses, but I didn't see a reference to a design to 'explain'.
====
>Thats not a Seas (although good guess, since Seas licensed the tech from KEF), it a KEF Ci200QR. Much closer to a real point source (although not a true one) than a duplexed array IMHO. Again hopefully GM can explain, since I'm not clear on how it all works and could be dead wrong.
====
Guess it's my day to wear the 'dunce cap'. What needs 'explaining'?

GM
 

AJinFLA

Banned
2005-02-09 4:35 am
Tampa
Hi GM,

sorry if I was not so clear (not unusual for me ;)), I was having difficulty understanding how the duplex approximates a point source. Both the Legacy (see pic above) and Donald's :
http://www.dnaudio.com/DNA Speaker.html
use multiple mids (phase?) arrayed around a tweeter. Donald's is probably closer to what you were describing, with a WG dome centered. He has rotated the mids axis slightly. Why? It is I who is the dunce, since I can't figure exactly how this all works. How does it approximate a point source with multiple drivers? Does the interference between the outputs cause cancellations off axis, to focus or beam the output towards the listening area, minimizing room effect, etc? Enquiring minds want to know. Neither Legacy nor DNA go into much detail about exactly how this is accomplished, since both are commercial offerings.

Cheers,

AJ
 
Greets!

Well, a typical 'fullrange' driver has several mechanical XOs that show up as notches on an unsmoothed impedance plot so it has a polar response that changes with frequency, but overall is more in time/smoother than a typical multi-way speaker with the same BW. A duplex moves the separate HF into the center of the polar field and when done right can have a better polar response and similar enough timing that it 'sounds' like a single driver.

WRT the DNAs, maybe I'm missing some subtlety, but let's view it as two MTMs with a common 'T'. If you search around and find a polar response of a MTM, visualize another one superimposed horizontally with the understanding that anywhere they overlap they will sum, so more gain, deeper nulls. Acoustically it will be uneven, creating a flower petal like pattern, so ideally two more MTMs are needed to average it all out, but then either the drivers have to be smaller or the XO point lower, so rotating them around a bit to find the best blend to the LF system seems as good a solution as any.

I've never experimented with it per se, but there's probably some psycho-acoustic trickery going on also since our hearing is horizontally weighted, so shifting them slightly may make it sound more uniformly 'round'. My SWAG is a ~22.5 deg offset.

Anyone up for some experimenting? ;)

GM
 
Didn´t you read the patent??

AJinFLA said:
He has rotated the mids axis slightly. Why?

AJ,
sorry I didn´t mention that you should read down the thread. So perhaps you didn´t read the patent, where it says: "... the centers of the four transducers forming a pattern wherein the centers of the four transducers coincide with corners of a square pattern where a bottom side of the square pattern forms an angle within the range of 18.4 to 31.0 degrees with respect to the horizontal axis to produce a coherent sound field, which reduces off-axis horizontal and off-axis vertical energy lobes".
 

AJinFLA

Banned
2005-02-09 4:35 am
Tampa
Hi Rudolf,

You are correct, I saw the HT guide thread the day after my questions to Donald on the Madisound board, but saw no new info being offered. I speed read through the thread (as I often do) at a later date and completely missed that Dennis H had kindly looked up the patent (something I would never have thought of doing :xeye: ). It is the type of design that is beyond the scope of what I can, or want, to do. How to design, measure (accurately) and assimilate the necessary data for a successful design like that is beyond my current capabilities. I can see its merits, but also its shortfalls. All designs are of course compromises. That particular approach doesn't seem quite right to me (personally). I have my own ideas about a HE, point source dipole and hopefully will have it ready so it can be judged by my peers at the next local DIY gathering.

Cheers,

AJ
 
I'm the Dennis H over at HTGuide. ;) My take on rotating the drivers is it's a bit of "specsmanship." Yes, it gives equal spacing of the drivers, and presumably smoother response, on the horizontal and vertical axes. But that only really helps because most people measuring the response will measure on the horizontal and vertical axes. But the real room power response is a sum of all axes so any difference in the sound by rotating them should be minimal at best.