LineSource said:
The SRO mid is a very high energy guitar driver and I put a jensen flexair surround on it to get it to work lower on the baffle and to eliminate (almost all of it) breakup My baffle will not allow it to go below 100 so I don't want a hole in the response there - it is like dead flat now crossing at 120 4th order. So yes, the modified SRO (and most 12's) will go to 80 Hz and cover the voice (at lower power handling) but not on my baffle.
I like the amp hooked to the SRO with 6" of wire - please don't make me any mmore paranoid then I already em
swak said:
The EV has the quality of the Lil Buddy but even better tone plus it is more efficient and will go lower, (and higher) with a nearly flat (like plus/minus 1db) from 120 to 2K. It is more 3D vivid (and more accurate!) in the midrange then the little super quads I have here - they are for sale too
This metal one, XT1086, is a lot cooler looking in an open baffle too and goes lower than the smaller 18 sound, A bit more money but not excessive
http://www.loudspeakersplus.com/product.asp?specific=1602
they claim to 1200hz
1500hz with constant vert dispersion
and 1400 with constant horiz. dispersion
http://www.eighteensound.com/index.aspx?mainMenu=view_product&pid=178
Ordered Altec 288-16H
Lynn,
I included the description that you wrote in my order:
"One pair GPA 288-16H compression drivers with low-power aluminum 16- ohm diaphragm, Tangerine phase plug, Alnico magnet, and no bug screens"
Is there anything else that I need to specify and let Bill of Great Plains know?
Lynn,
I included the description that you wrote in my order:
"One pair GPA 288-16H compression drivers with low-power aluminum 16- ohm diaphragm, Tangerine phase plug, Alnico magnet, and no bug screens"
Is there anything else that I need to specify and let Bill of Great Plains know?
agent.5 said:
That is the complete description, and you will get the same drivers I'm getting. All you need after that are horns or waveguides that match the 1.4" exit and preferably also match the internal 8-degree, 207 Hz flare of the compression driver. At the bottom of the post, you'll see the mold that Martin is building for the AH-425. Nice autumn weather in Perth, I see.
pdan said:
Not quite as arbitrary as that. I've never been a fan of the extreme HF of any horn or waveguide, and now that I've gone down the rabbit hole of horn/waveguide design, I have a better idea why. Once the wavelengths start to approach the size of the phase plug, many things go wrong:
1) The wavefront coming out of the phase-plug gets rough thanks to non-summation from the multiple exits - the same problem as line-sources on a much smaller scale.
2) Generation of HOM's gets more severe with higher frequencies (multipath bouncing around inside the horn), which is made worse by the non-coherent wavefront emerging from the phase plug.
3) Distortion from the diaphragm goes up due to ...
4) The gain from the horn/waveguide going down, so that at the highest frequencies there is effectively no horn-gain at all.
5) Diaphragm breakup makes its appearance. Even small-format compression drivers have significantly larger diaphragms than comparable direct-radiators, so diaphragm breakup appears at frequencies 1.5 to 3 times lower than equivalent 1" direct-radiator domes.
6) HF boost equalization in the same frequency range where all of the above effects are happening at once, making them even worse.
So there are things going in the extreme HF of horns and waveguides that don't happen with conventional direct-radiator domes or ribbons. True, these lower-efficiency devices have their own set of problems, mostly with power-handling and lack of headroom, but the impulse response and CSD (of the best examples) is generally superior to horns and waveguides.
Put another way, I'd take a ribbon over a TAD or JBL beryllium compression driver any day of the week - above 10 kHz, that is. The short-wavelength world of the top octave is completely different than the 1~3 kHz world in terms of the best solution for the problem. The power levels are far lower, so low-mass diaphragms with simple, unobstructed emission paths (no phase plugs) are more attractive. This is inherent in the technology.
It is possible to build high-performance horn supertweeters (with smooth response to 30~50 kHz), but I still prefer a ribbon to a Goto or Ale - I've heard them and thought the Japanese exotics were OK, but nothing all that special.
I prefer a driver-centric, bottom-up approach to designing speakers. Other folks take a top-down, an idealized-speaker-should-do-this approach. That's never worked for me, although it works for others. I look at the best available technology (by measurement and audition) for a given portion of the spectrum, and design around that. In terms of measurement and audition, I feel that direct-radiators are the best choice above 10 kHz.
In the 1~10 kHz region, compression drivers play to their strengths, although impulse response with rapid decay characteristics remains the biggest challenge. In the region from the bottom of the spectrum to 300 Hz, direct-radiator woofers are the best choice for anyone who isn't willing to build a garage-sized bass horn. In the 300 Hz to 3 kHz region, direct-radiators and horns are slugging it out, with tradeoffs between distortion, impulse response, headroom, and overall size.
Attachments
Mr Olson,
Thank you for your post, it was very much appreciated.
I was under the impression though, that the RAAL was (is) the White Rabbit that lead you down into the horn labyrinth in the first place, as only compression driven horns are judged to be a suitable sonic match. So it seemed to me that the design has become a matter of the tail wagging the dog.
I must confess a strong prejudice - I can't stand the sound of compression driven horns. Well, at least if Advantgarde loudspeakers are anything to go by; even at low levels, driven by SETs, they sounded horrid. Later, when I found out how compression drivers worked I was mystified how anyone could take them seriously:
A sealed felt-lined back chamber that is supposed to absorb the back wave??
A metal dome diaphragm... ugh!
An area of high pressure coupled to radial slots or holes- coupled to a trumpet ???
I know that I've much to learn, in particular: prejudice is not a valid ground for aesthetic judgments. But ... ?
When the compression horn is up 'n' running, I'd be very interested in Mrs Olsons judgment. Who knows, perhaps she'll be your Ariadne ?
Cilla
Thank you for your post, it was very much appreciated.
I was under the impression though, that the RAAL was (is) the White Rabbit that lead you down into the horn labyrinth in the first place, as only compression driven horns are judged to be a suitable sonic match. So it seemed to me that the design has become a matter of the tail wagging the dog.
I must confess a strong prejudice - I can't stand the sound of compression driven horns. Well, at least if Advantgarde loudspeakers are anything to go by; even at low levels, driven by SETs, they sounded horrid. Later, when I found out how compression drivers worked I was mystified how anyone could take them seriously:
A sealed felt-lined back chamber that is supposed to absorb the back wave??
A metal dome diaphragm... ugh!
An area of high pressure coupled to radial slots or holes- coupled to a trumpet ???
I know that I've much to learn, in particular: prejudice is not a valid ground for aesthetic judgments. But ... ?
When the compression horn is up 'n' running, I'd be very interested in Mrs Olsons judgment. Who knows, perhaps she'll be your Ariadne ?
Cilla
Hello Mr Olson,
I'd like to point out something that looks to be a great 12" midrange driver. Look at the frequency graph!
12" midrange
I guess this 21 " neo driver would suit well bellow:
By the way how would you interpret the impedance curve of the 12 incher?
Thank you!
I'd like to point out something that looks to be a great 12" midrange driver. Look at the frequency graph!
12" midrange
I guess this 21 " neo driver would suit well bellow:
By the way how would you interpret the impedance curve of the 12 incher?
Thank you!
Some other aspect to be mentioned. The driver has a 2.5" voice coil. Probably if the cone would have been lighter, about 25-30 grams, instead of 50, it would have hit 103dB. Probably at the price of a not so well controlled breakup. It has a Bl of 22 T/m, that's something...
Maybe the motor of the Precision Devices driver, the cone of the 12NDA520 and AIC could make up THE MIDRANGE driver? Well.. you can't have everything...
The 21 also looks very well behaved.
Maybe the motor of the Precision Devices driver, the cone of the 12NDA520 and AIC could make up THE MIDRANGE driver? Well.. you can't have everything...
The 21 also looks very well behaved.
SunRa said:
There are definitely some goofy things going on with the 12" driver. You can see a bunch of bumps in the impedance curve that are due to one of 2 things. First, the driver was moving when the curve was taken making it pretty much invalid all together. There also isn't a nice clean peak, it's more rounded and not even, again making me think the driver wasn't properly secured. The other thing is that every one of those bumps is due to some kind of resonance. Which looking at the response curve I'd guess some of them are. Just under 3KHz there is a bump in the curve corresponding to a dip in the response curve for example. Most of the bumps in the impedance curve have something that seems to correspond on the frequency response as well. To really tell what's going on you need to see both on and off axis and ideally a CSD to tell if there are resonance issues associated with the bumps in impedance curve.
Sunra mentions the BL of 22 being high. Again as Nick has pointed out, BL without RE means nothing as they are in proportion to one another. BL^2/Re is about the best way to compare motor strength between drivers. In this case the BL^2/Re is 84. Fairly high motor strength, but nothing that really stands out. Our standard 2.5" coil driver with 23mm Xmax has BL^2/Re of 142 after taking off steel on the pole to fit the copper sleeve. Without the copper sleeve and more steel on the pole it's in the range of 160. In reality though, none of this makes a whole lot of difference. It is the correct ration of motor strength, mass, and compliance that all has to be considered for the application.
John
Thank you for your response John, I realise now that I did the mistake to link this driver again after I did it a couple of months ago and Mr. Olson did the same comment regarding the reflection of the impedance bumps in the frequency curve.
I am very much interested in what it takes to make a high spl (103db or more) 12 driver. The ones I know about are only the supravox 12" electromagnet version and the EV SRO Magnetar is using.
I think that a major design point is a strong motor and probably low mass. I don't understand very well why Bl/Re ratio is more significant, could you please help me with a link to Nick's comments on this? And where could I find a good discussion regarding the compliance of a driver?
I hope this is not too much off-topic.
Thank you!
I am very much interested in what it takes to make a high spl (103db or more) 12 driver. The ones I know about are only the supravox 12" electromagnet version and the EV SRO Magnetar is using.
I think that a major design point is a strong motor and probably low mass. I don't understand very well why Bl/Re ratio is more significant, could you please help me with a link to Nick's comments on this? And where could I find a good discussion regarding the compliance of a driver?
I hope this is not too much off-topic.
Thank you!
SunRa said:
Here is a good example, lets take:
BL = 14.5, BL = 17, BL = 20.1
Which is the strongest motor?
...
If you guess they are all exactly the same motor then you would be correct. These are all from TD series drivers with a Lambda phase plug motor.
Next question which one comes from the driver with the highest sensitivity?
...
If you would have guessed the middle spec of 17 then you would be correct.
Now figuring with Bl^2/RE we get:
First spec is from a 4ohm version of the TDxH series = 58.4
Middle spec is from a 8 ohm version of the TDxM series = 43.8
Biggest spec is from a 8 ohm version of the TDxH series = 62.2
As you can see the most efficient driver is also the "weakest" motor. Our motor design is such that a large amount of flux is purposely "leaked" outside the gap to increase the linear xmax. The TDxM series with their shorter coils simply do not intersect as much flux lines at the same time as the longer coil versions. Hence their relative "motor strengths" are in fact weaker. This is akin to how an underhung motor wastes the majority of its flux lines to gain highly linear xmax. The TDxM series gain the extra sensitivity over the other TD series all by the lower moving mass. The difference between the 4ohm and 8ohm versions of the same driver listed here is due to tolerance variance and also the winding factor in the coil. You can design one size of wire for maximum efficiency with a fixed coil gap and wind height, everything else is compromised slightly.
Clear as mud right
nickmckinney said:
The TDxM series with their shorter coils simply do not intersect as much flux lines at the same time as the longer coil versions. Hence their relative "motor strengths" are in fact weaker. This is akin to how an underhung motor wastes the majority of its flux lines to gain highly linear xmax.
Clear as mud right
Hmm, seems a bit "muddy".
That underhung maybe p!ssing away the total motor's field strength.. BUT it will likely exhibit greater "strength"/control over the vc's very short excursion (which is often less than what is described as "linear"). To compensate many pro manufacturer's increase the vc's gauge and diameter.. but that increases mass, and unfortunately they also tend to make the driver less compliant in an effort to keep the vc from leaving the gap (often under *harsh* operation). One step forward.. two steps back (..well, for sound quality rather than durability).
pdan said:
No, the RAAL was never the "White Rabbit" that led to everything else.
To put things in a larger perspective, I designed the time-aligned LO-2 and associated push-pull subwoofer with a Bessel alignment in 1979 and kept it, until I finished the Ariel in 1993. The Ariel was clearly better, so I retired the LO-2. It's now 2007~2008, and I've outgrown the Ariel. When I'm satisfied, I stay with the result for a long time, and turn my attention to other parts of audio.
The only reason I design anything is to resolve certain problems that have not been solved by the commercial sector. At the time of the LO-2, the time-aligned satellites were 6 months ahead of the Spica TC-50, and the push-pull Bessel-alignment subwoofer was a decade or more ahead any commercial equivalent. The Ariel was one of the first narrow-profile MTM transmission lines with medium efficiency (92 dB, still higher than many audiophile speakers).
Bud Purvine encouraged me to design my next speaker in a public forum, instead of the private consultations I usually use during the development phase. It was a good way to get beyond the intense self-pity and emotional turmoil of the recovery from the accident, so I took him up on it - thus, this thread you see here.
My initial naive hope was that a 12~15" coax in an open baffle could be combined with a bass array to get the results I wanted. Alas, every single coax not only had pretty rough response from the horn (which was expected) but rough response from the woofer cone as well. Woofer cones don't seem to like having a big open internal space in the middle, it seems, nor horns obstructing the emissive surface. The published "smooth" curves of coaxes seems to be a lot of smoke and mirrors - 1/3 octave curve smoothing combined with each driver more or less filling the rough spots of the other. The curves of the individual drivers was not at all attractive.
The other data point was that I have heard - very rarely - compression drivers and horns sound rather good. Once was at the San Francisco show a few years ago, when I took the Karna amplifier along with Gary Pimm demo'ing his amplifier. The Altec A5's, although extremely crude in many ways, actually had pretty nice mids and highs, and effortlessly filled the 300-seat auditorium. That was with a 288 compression driver and the big 1505 multicell horn, which sounded nothing at all like the garbage mids and highs from an ordinary Altec A7, which uses the small-format driver and sectoral horn.
I've also heard not one, but three systems in the Denver area that combined the Oris, and later, the Azura front horn with Lowther and AER drivers. Although there was more coloration than I care for, these systems did a lot of things very right, and I agreed with John Atwood - who has pretty good taste in these things - that the sound of the best of these was "world class". One of these systems replaced a fully tricked-out Avante-Garde Trio, and was superior to the AG in every single way. (Karna has heard all three of these systems, liked their virtues, but wanted lower levels of coloration. Agreed.)
So I've had several "example proofs" that horns don't have to sound like complete garbage - that in fact, they can lose almost all of the horn coloration, and gain some very unusual qualities that just aren't there with any direct radiator. Effortless dynamics, vivid tonality, and a sense of "presence" and immediacy that I ascribe to much lower IM distortion and absence of voice-coil heating effects.
Traditional horns have appalling time-domain response, far worse than conventional direct-radiators. The worst of all seem to the ones used in sound reinforcement and theaters for the last twenty years - descendants of the Altec Manta-Ray and JBL Bi-Radial horns. The older "sectoral" horns, as used in the Altec A7, are really bad as well, along with all the horns that use horn-lenses and other diffractive tricks to control dispersion.
Diffraction is undesirable in direct-radiator cabinets, where it can obscure image quality and curtail the depth impression. It is far worse in horns, where it results in gross "PA" colorations - which are still very much there in modern high-end horn systems.
As far as I can tell, there are only three horn or waveguides that appear to have low diffraction: the Dr. Geddes-designed OS, the Le Cleac'h, and the Dr. Holland-designed AX-2. The real proof is in the measurements, and I haven't seen time-domain measurements of the Le Cleac'h and the AX-2. The time-domain measurements of the Dr. Geddes OS that have appeared in this forum have been impressive, the best I've seen for any horn or waveguide. I look forward to measuring the new AH-425 & 288-16H on the MLSSA system. No data on the proprietary AX-2, so no real ideas there.
In terms of commercially available horn systems that have low coloration, I've only heard two: the Geddes Summa and the AudioKinesis. I haven't heard the JBL K2 9800, but the shape of the horn does not inspire confidence - the termination of the mouth-edge looks pretty sharp to me, and the design problems of a horn that is substantially wider than it is high are not trivial to solve. I'd still like to hear the JBL K2 9800, since it is a "flagship" product designed by talented engineers with long experience in the professional sector.
The combination of the San Francisco experience with the 288 & 1505, three world-class systems in the local area, and the audition of the Summa/Ai & AudioKinesis tells me that horns and waveguides can have low coloration as well as dynamics that are beyond the reach of low-to-medium efficiency direct-radiators. If they can do it, I think I have at least a reasonable chance to get there as well - at least I know it is possible.
Unfortunately, from what I've heard at shows, dealerships, and in various homes, almost every $15,000 to $150,000 audiophile horn system is a disaster, matched only by the terrible sound of mainstream (low-efficiency) audiophile speakers in the same price range. HF screech, midrange harshness and outright cone breakup, combined with overall gross levels of coloration seem to be the order of the day.
I just don't understand, except to acknowledge that the boutique magazines and websites seem to be actively encouraging this state of affairs. In terms of the mainstream high-end, there are only handful of low-coloration products available, and with the exception of the Summa/Ai and AudioKinesis, I know of none that are high-efficiency.
ScottG said:
Hmm, seems a bit "muddy".
That underhung maybe p!ssing away the total motor's field strength.. BUT it will likely exhibit greater "strength"/control over the vc's very short excursion (which is often less than what is described as "linear"). To compensate many pro manufacturer's increase the vc's gauge and diameter.. but that increases mass, and unfortunately they also tend to make the driver less compliant in an effort to keep the vc from leaving the gap (often under *harsh* operation). One step forward.. two steps back (..well, for sound quality rather than durability).
Very true, as I am fond of saying "everything in a speaker is a compromise, please choose yours"
Lynn Olson said:
What are your thoughts on the Danley Unity? That 60 x 60 lens with a compression driver only is some of the sweetest 1-10Khz I have ever heard myself.
Lynn Olson said:
And for this we are thankful. For those of us (like myself) that can't add a single word of knowledge or advice on the subject, it is a good learning experience. It helps us appreciate what degree of work, analysis, and research goes into a properly designed speaker system.