Accuton vs Scanspeak vs Audio Technology

[JerryLove]

You know you pointed both links at a woofer?

 

[TLS Guy]

This post is way off track.

Docks, this is your first project.

You want a good bookshelf, with good power output, that will mate well with a sub and integrate easily. Right?

You want a first project with a high chance of success.

You absolutely do not need to make this a three way no what another poster insists.

Good speakers use the least number of crossovers possible and try and do the least damage with every one. This makes the acoustic responses of your drivers and the T/S parameters of the woofer paramount. What the drivers are made of and there mode of operation is VERY secondary.

People who drone on about the properties of woofer cone materials etc. have never designed and built a speaker that was any good.

I have selected for you a woofer and tweeter that will mate well together and give you good power handling as a THX type arrangement crossing over to the sub at 80 Hz.

Your woofer.

Your tweeter.

You can crossover very happily in the 2 to 2.5 kHz range.

I have not played with the crossovers, but something like fourth order for the woofer and second order for the tweeter will likely give you a nice composite fourth order response.

However making a good crossover is the real art. I suspect that the woofer crossover may have to change orders.

I think these drivers with a good crossover and a sealed alignment is capable of very good results indeed.

If you want higher power handling and a better horizontal response with limitation of the vertical response, these drivers would make a very nice small sealed MTM.

Thanks Jerry!

This is the tweeter.

The woofer.

 

[GranteedEV]

I'm not really following the sentiments in this thread. The RAAL 70-20XR is an extremely robust tweeter used in plenty of excellent speakers, 2-way or 3-way. It can be crossed over near 2khz. The Salk Soundscapes, by many considered very dynamic speakers, have it crossed at about 1.9khz.

Is it probably best used in a 3-way for a smooth sound power response? Yes. But that's not something dictated by the ribbon's limitations. Certainly I doubt the designers at Vapor Sound, Salk Sound, and Ascend Acoustics feel that way.

Obtaining the RAAL 70-20XR is a non issue for me. All this talk about the downfalls has me reconsidering it now...

Unless those people have actually messed with a 70-20XR, I don't think they're in a useful position for you. Of course that also means you may not be in a useful position to make that decision either. That's the crux of it when we're talking about a $300+ tweeter. Many are perfectly happy with $50 soft dome tweeters like the Dayton RS28F and can't hear any useful difference.

If i didn't take this route what dome tweeters would you suggest? First thing that comes to mind is the Scanspeak 6600

If I were doing a speaker with a dome tweeter, I would use:

This + #7125475125

As far as measurements of woofers go, I don't much subscribe to "Zaph" measurements of harmonic distortion etc. They can be useful but I'm not sure they actually correlate to what you hear. I do look at Klippel mesurements, which favor drivers like the Scanspeak Illuminator 18WU, Exodus EX-Anarchy, and Acoustic Elegance TD6H. W/M Efficiency matters too. As do other factors. The nice thing about some of these Phase Plug drivers like the Seas W18 is that heat is conducted out of the box rather than in (which is what a pole piece vent will do). Keeping heat low can be pretty valuable in a powerful setup.

There's really a ton of options out there, and within their piston and linear throw range there might not be major differences. Some drivers optimized for long excursions DO seem to have some abberations though, seen in the frequency response. This can include high end Scanspeak Revelators.

If I'm not mistaken you have some killer subwoofers and i doubt any 6 inch woofer can keep up in output. It's certainly possible that for your goals, a small two-way is not what you're looking for. I can't answer that question for you. There's a lot of factors at play and some of us would advise you to build some experimental designs first before going for the gusto with $300 drivers.

See what you yourself can do with some good $50-75 drivers - IE Dayton RS28F tweeter and Aurum Cantus AC165. Learn what you personally want or don't want before you gamble on some exotic drivers. Your room might dictate a different polar pattern (wider, narrower, smoother in the midrange more vertically restricted, etc). You might need more headroom. Play around with a full range driver too, just to hear the difference between a crossover and lack thereof.

 

[Dennis Murphy]

It always depresses me to see posts slamming ribbon tweeters based on erroneous preconceptions and lack of actual listening experience. I can use basically any tweeter--dome or ribbon--that's out there, but I have settled on the RAAL ribbons simply because there isn't anything smoother, more realistic, or easier to work with. They are not at all difficult to integrate with a woofer or midrange. In addition, I've made countless measurements of both the OEM 20mm and the more available 10 mm, and there are no diffraction effects from the sawtooth baffle opening. None. The only downside of the ribbons is that you are pretty much restricted to a 4th order acoustic slope if you want to cross fairly low in their operating range. That's not a bid deal for me, because 2nd order slopes are very hard to implement with any tweeter unless you have a sloped or set-back baffle for them. Of course, none of this should be construed to mean that speakers with quality dome tweeters aren't capable of superb performance. I've done many designs with ScanSpeak, Seas Excel, Morel, and many less expensive domes that were very accurate and enjoyable. I just happen to find the RAAL sound more natural and open at the top, and with less of a ssssssssssssssssssss quality, than any of the domes I've tried.

 

[DS-21]

It always depresses me to see posts slamming ribbon tweeters based on erroneous preconceptions and lack of actual listening experience. I can use basically any tweeter--dome or ribbon--that's out there, but I have settled on the RAAL ribbons simply because there isn't anything smoother, more realistic, or easier to work with. They are not at all difficult to integrate with a woofer or midrange. In addition, I've made countless measurements of both the OEM 20mm and the more available 10 mm, and there are no diffraction effects from the sawtooth baffle opening. None. The only downside of the ribbons is that you are pretty much restricted to a 4th order acoustic slope if you want to cross fairly low in their operating range. That's not a bid deal for me, because 2nd order slopes are very hard to implement with any tweeter unless you have a sloped or set-back baffle for them. Of course, none of this should be construed to mean that speakers with quality dome tweeters aren't capable of superb performance. I've done many designs with ScanSpeak, Seas Excel, Morel, and many less expensive domes that were very accurate and enjoyable. I just happen to find the RAAL sound more natural and open at the top, and with less of a ssssssssssssssssssss quality, than any of the domes I've tried.

If one uses a 10mm wide tweeter firing through a diffraction slot on a 180 waveguide (i.e. a flat baffle)...I cannot envision that sounding non-awful. Physics dictate that will have a massive directivity shift in the midrange.

Now, with a 3" wide mid and a low crossover, that's a different ballgame if done right.

 

[Dennis Murphy]

If one uses a 10mm wide tweeter firing through a diffraction slot on a 180 waveguide (i.e. a flat baffle)...I cannot envision that sounding non-awful. Physics dictate that will have a massive directivity shift in the midrange.

Now, with a 3" wide mid and a low crossover, that's a different ballgame if done right.

What do you define as a low crossover? The RAAL 10 was designed to be crossed at 2.9 kHz. Here is the response for the Philharmonic 3 at 60 degrees off axis. In a 2-way crossed to a 6" woofer at 2.9 kHz, there would be a shift in directivity, as there would be if it were crosssed to a dome tweeter at that frequency.

 

[DS-21]

What do you define as a low crossover?

Where the mid is still radiating more or less hemispherically. Anything else, with a tweeter loaded on a 180deg waveguide, is just poor design.

 

[Dennis Murphy]

Where the mid is still radiating more or less hemispherically. Anything else, with a tweeter loaded on a 180deg waveguide, is just poor design.

I'm really not sure what your point is. There are a bajillion 2-way designs with dome tweeters crossed above the point of "hemispherical" radiation for the woofer. The vast majority of 2-way designs fit that description. This has nothing to do with with domes vs. ribbons. And if the crossover is decent, these designs will sound fine. Not perfect--there will be a disconnect in the radiation pattern of the woofer and tweeter. But nothing that constitutes inherently "poor design."

 

[Swerd]

If one uses a 10mm wide tweeter firing through a diffraction slot on a 180 waveguide (i.e. a flat baffle)...I cannot envision that sounding non-awful. Physics dictate that will have a massive directivity shift in the midrange.

Physics shows that it should work well, without any "massive directivity shift".

Diffraction interference patterns occur whenever a slit with two edges is encountered by waves where the wavelength is roughly similar to the dimensions of the slit (paraphrased from wikipedia).

Lower down on that same wikipedia page, there are some useful gif animations that model diffraction patterns. Scroll down to where it says Diffraction of Light and see the two green gifs on the right.

The first gif shows the resulting pattern when a slit width is equal to the wavelength of an incident plane wave. There is little or no interference.

The second one shows the diffraction interference pattern from a slit width 5 times larger than the wavelength

Sound at 2,000 Hz has a wavelength of 172 mm. That's 17.2 times larger than the 10 mm slit width, and 8.6 times larger than the 20 mm slit. No diffraction can happen at 2,000 Hz because the ribbon slit width is invisible to sound at 2,000 Hz.

At 20,000 Hz the wavelength is 17.2 mm, roughly equal to either 10 or 20 mm (approximately what is shown in the first gif animation), but not at all close to a slit width 5× the wavelength, as illustrated in the second gif.

An opening in a ribbon tweeter with a slit 10 mm or 20 mm wide cannot generate diffraction interference with sound at 172 mm wavelength (2,000 Hz). At 17.2 mm (20,000 Hz) it is just beginning to happen, but no one has demonstrated whether its affects are measureable or audible.

The sawtooth pattern of the opening of the RAAL tweeters varies the actual dimensions of the slit opening. Compared to openings with straight, parallel edges, I would guess that the saw tooth pattern has the effect of minimizing any interference patterns due to diffraction.

 

[Steve81]

I'm really not sure what your point is. There are a bajillion 2-way designs with dome tweeters crossed above the point of "hemispherical" radiation for the woofer. The vast majority of 2-way designs fit that description. This has nothing to do with with domes vs. ribbons. And if the crossover is decent, these designs will sound fine. Not perfect--there will be a disconnect in the radiation pattern of the woofer and tweeter. But nothing that constitutes inherently "poor design."

If I might offer an opinion, I would think my bookshelves versus the Philharmonics would be a good representative case of this. By DS-21s standards, my speakers are poorly designed; certainly I'd agree they aren't perfection, and there is no question that the Philharmonics cleaned their clock in terms of having a much fuller presentation (even discounting the differences in bass extension), a good part of that being attributable to the much wider and more even dispersion of the Phils. Still, I didn't walk away thinking that I needed to toss my speakers in the garbage, nor did the thought cross my mind that they are a horrible design; to me, it's more a matter of compromising to meet a price point and certain performance objectives.

 

[Swerd]

I could have worded some things better:

The first gif shows the resulting pattern when a slit width is equal to the wavelength of an incident plane wave. There is little or no interference.

When the slit width is equal to or greater than the wavelength of the incident plane wave, there is little or no interference. The waves passing through the slit act as if they came from a single point source.

The second one shows the diffraction interference pattern from a slit width 5 times larger than the wavelength

When the slit width is much longer than the wavelength, each edge of the slit acts as if its a point source of sound. Because these two point sources are close enough together, the waves coming from them create interference patterns that also occur when two drivers are close to each other.

Sound at 2,000 Hz has a wavelength of 172 mm. That's 17.2 times larger than the 10 mm slit width, and 8.6 times larger than the 20 mm slit. No diffraction can happen at 2,000 Hz because the ribbon slit width is invisible to sound at 2,000 Hz.

No diffraction can happen at 2,000 Hz or lower because the ribbon slit width is so small compared to the sound waves that the waves act as if they came from a single point source.

 

[GranteedEV]

I can't believe Swerd is this concerned about diffraction a day after the hurricane - I take it you made it out in one piece!

 

[Swerd]

I can't believe Swerd is this concerned about diffraction a day after the hurricane - I take it you made it out in one piece!

Yep, all in one piece . My electricity flickered a bunch of times, but I never lost power. Today it's 42° F and raining, so central heat is very much appreciated

I went in to work yesterday, and left shortly after noon because the wind was noticeably picking up. The roads were nearly empty but with high winds and heavy rain, anything can happen. I brought work home so today I can say "I worked at home" .

When I read that BS about diffraction and ribbon tweeters, it seemed wrong. But it took me a while to remember basic optics. While waiting out Sandy, I had plenty of time and eventually found the answer. I'm relieved to know that the basic physics of wave behavior haven't changed since I learned it last century.

 

[DS-21]

I'm really not sure what your point is. There are a bajillion 2-way designs with dome tweeters crossed above the point of "hemispherical" radiation for the woofer.

So what? That most commercial and DIY loudspeakers made have poor midrange performance and thus are simply not high fidelity devices is neither a new nor an interesting observation at this point.

The vast majority of 2-way designs fit that description. This has nothing to do with with domes vs. ribbons. And if the crossover is decent, these designs will sound fine. Not perfect--there will be a disconnect in the radiation pattern of the woofer and tweeter. But nothing that constitutes inherently "poor design."

That disconnect is the inherently poor design. A competently designed loudspeaker cannot have a directivity shift in the midrange. If it does, it's simply a crappy speaker.

It's telling that on the speakers you sell under your own label, you use not only a very narrow midrange, but also a low mid-tweet crossover.

Physics shows that it should work well, without any "massive directivity shift".

The ellipses you put in my quote is, of course, the part that actually mattered.

Sound at 2,000 Hz has a wavelength of 172 mm. That's 17.2 times larger than the 10 mm slit width, and 8.6 times larger than the 20 mm slit. No diffraction can happen at 2,000 Hz because the ribbon slit width is invisible to sound at 2,000 Hz.

You're conflating two concepts here that could not be analytically more distinct concepts:

1) The midrange mushroom cloud that happens when you cross a beaming woofer to an unconstrained tweeter, creating a speaker that, by (lack of) design is doomed to be a low-fidelity device no matter how brilliantly engineered the crossover may be.

2) Listening fatigue from diffraction/HOM's. Compare the dimensions of the diffraction slot in those Reynolds Wrap fetish objects with the diffraction slot in an old "CD" horn, and get back to us.

 

[Nuance AH]

Dennis,

DS-21 is a shill of waveguide only designs. That's fine, and everyone has their opinions, but then he goes around and slams every other design and/or tweeter without even hearing them. He looks at pictures and determines if a speaker is crappy sounding or not. He's very full of himself and it is well known he's a troll and should just be ignored. You're far more knowledgeable and experienced than he, so it probably isn't worth your time debating with him, as he'll never admit he is wrong (and he IS wrong); he'll just drag you down to his level of foolishness and beat you with experience.

Just ignore him, Dennis. The people who actually know what's going on know your'e one of the best crossover designers on the planet, and we aren't questioning your experience and/or skills. They also know not to take DS-21 seriously. Personally I think he's jealous (which of course he'll deny).

Yes, I realize I'll get called out for slinging mud, but someone has to stand up and call a spade a spade. 'Nuff said.

 

[Dennis Murphy]

So what? That most commercial and DIY loudspeakers made have poor midrange performance and thus are simply not high fidelity devices is neither a new nor an interesting observation at this point.



That disconnect is the inherently poor design. A competently designed loudspeaker cannot have a directivity shift in the midrange. If it does, it's simply a crappy speaker.

It's telling that on the speakers you sell under your own label, you use not only a very narrow midrange, but also a low mid-tweet crossover.



The ellipses you put in my quote is, of course, the part that actually mattered.



You're conflating two concepts here that could not be analytically more distinct concepts:

1) The midrange mushroom cloud that happens when you cross a beaming woofer to an unconstrained tweeter, creating a speaker that, by (lack of) design is doomed to be a low-fidelity device no matter how brilliantly engineered the crossover may be.

2) Listening fatigue from diffraction/HOM's. Compare the dimensions of the diffraction slot in those Reynolds Wrap fetish objects with the diffraction slot in an old "CD" horn, and get back to us.

You are completely hung up on one aspect of speaker design. However, you did get me curious enough about speakers that are specifrically designed for CD that I arranged to borrow an expensive and much-ballyhooed specimen. I measured to confirm that the radiation pattern was indeed uniform off-axis, which it was. That's about all I did with it, because I simply couldn't listen to the colored midrange and strident highs. It's sitting about 5 feet from me in all of its bulk and heaviness, and I'll have to figure out a way to get it upstairs for return shipping. I'm sure there are good CD speakers out there, as there are many good non-CD speakers. But one thing for sure, I've never heard midrange coloration like that in a quality conventional 2-way speaker even when crossed where the woofer is starting to beam. Those speakers certainly fall short of perfection. But all speakers do, and to slam one category of speakers simply because they aren't perfect in one design aspect reflects a lack of understanding of the speaker design process. But enough--I don't think anyone else is interested in this discussion.

 

[Swerd]

The ellipses you put in my quote is, of course, the part that actually mattered.

The ellipses in your quote were your own. See below for that post quoted in full.

If one uses a 10mm wide tweeter firing through a diffraction slot on a 180 waveguide (i.e. a flat baffle)...I cannot envision that sounding non-awful. Physics dictate that will have a massive directivity shift in the midrange.

Now, with a 3" wide mid and a low crossover, that's a different ballgame if done right.

I am capable of respecting your opinions about speaker design, even if they are different. It might help me (and others) to understand them if you at least tried to explain them. So far you haven't.

You have spoken of your opinions indirectly, only to the extent that you dismiss, as foolishly wrong, anyone who doesn't adhere to your ideas. Just what are those ideas, and why should we be interested in them? To be crystal clear, I'm saying that evidence might convince us of any merit in these ideas. Backhanded insults will not substitute.

But I will not accept crude attempts to put words into my mouth.

 

[GranteedEV]

However, you did get me curious enough about speakers that are specifrically designed for CD that I arranged to borrow an expensive and much-ballyhooed specimen. I measured to confirm that the radiation pattern was indeed uniform off-axis, which it was. That's about all I did with it, because I simply couldn't listen to the colored midrange and strident highs. It's sitting about 5 feet from me in all of its bulk and heaviness, and I'll have to figure out a way to get it upstairs for return shipping.

Why do you have to pique our curiousity with such vagueness?!?!

 

[ousooner2]

* taken from another thread

There are four main ways of which I know to competently design a loudspeaker:

1) A waveguide on the tweeter, chosen such that it matches the directivity of the next driver down in the desired crossover region. This results in a choice of pattern. This is what Revel, JBL, Behringer, Mackie, GedLee, David Smith-era Snell, and some others do. (Yes, the $300/pair Behringer B2031p is a better speaker than most $20,000 "high end" speakers.)

2) A concentric driver, using the cone as a waveguide to set the tweeter's pattern. Tannoy, KEF, GedLee, Soundfield Audio, and others use this approach. The Danley Synergy horn is a riff on this approach, too, that has some very interesting advantages compared to a moving-cone waveguide.

Note that not all concentrics work. Some are just generally unsuitable for music reproduction, such as the awful-sounding flat Thiel diffraction generator in the 3.7. (Anyone who thinks the Q900 sounds fatiguing over time - I'm one of 'em, mind - should try out the Thiel CS 3.7 to really hear fatiguing!) Some don't use a tweeter stout enough to play down to the frequencies needed to match directivity and thus negate most of the benefits of concentric tweeter mounting, such as the previous-generation KEF xQ/iQ speakers.

3) A very small midrange and a crossover low enough such that they're both more-or-less hemisphere radiators in the crossover region. This results in uniform, but very broad, coverage. Great for an absorptive room, not so good for a live room. Good commercial examples of this approach are the Philharmonic Audio 3-ways, the NaO Note, and some NHT's. The Linkwitz Orion is also mostly in this category, due to the very very low crossover between tweeter and midwoofer.

4) An electrostat with concentric delay rings such that the speaker radiates like a section of a sphere. Quad is the only one using this approach to my knowledge. The ESL-63 and successors are outstanding speakers, if you can live with wide panels well out into the room and a narrow listening area due to the very narrow treble dispersion.

Note also that using one of these four approaches far from guarantees a good loudspeaker. Think "necessity" not "sufficiency." Furthermore, something that passes the "eyeball test" may in fact not meet the criteria. The previous-generation iQ/xQ KEF's are a good example of that.



Listen for unnatural spittiness in the lower treble of live and unamplified music, that one simply wouldn't hear in a live performance. Preferably using an A/B comparison with a known competently-designed speaker and an incompetently-designed speaker. That's the easiest giveaway. The Brothers-in-Arms DVD-A isn't necessarily the best choice, because the unnecessary mixing of instruments into the surrounds is quite distracting. It's one of those disks I'm more likely to listen to in 2-channel-expanded-to-surround-via-DPL2 than discrete multichannel.

 

[monkish54]

He's very full of himself and it is well known he's a troll and should just be ignored. You're far more knowledgeable and experienced than he, so it probably isn't worth your time debating with him, as he'll never admit he is wrong (and he IS wrong); he'll just drag you down to his level of foolishness and beat you with experience.

While I don't agree with most of what DS has been saying in this thread, I don't think it is appropriate to accuse him of what you have here. If you want to refute his argument, that is one thing, but to attack his character is another.

The people who actually know what's going on know you're one of the best crossover designers on the planet, and we aren't questioning your experience and/or skills.

Dennis' skills have not been brought to question. They were simply disagreeing on a particular part of speaker design. I don't think anyone here thinks Dennis is a poor designer, but even if he was THE BEST, what's wrong with questioning or disagreeing? I greatly respect Dennis and have even asked him to design a crossover for me, but just because I think he is talented, doesn't mean he should be immune to questioning/explaining his reasoning. How would one learn if they never question?

If Émilie du Châtelet had thought Newton was too talented to question she would have never considered Gottfried Wilhelm Leibniz could be right, and we wouldn't have realized he was.

http://en.wikipedia.org/wiki/Vis_viva

https://www.youtube.com/watch?v=219YybX66MY&feature=related

I think he's jealous (which of course he'll deny).

Do you have any evidence for this?

Yes, I realize I'll get called out for slinging mud, but someone has to stand up and call a spade a spade.

It seems more like you are attacking his character in order to disprove his argument.