Beyond the Ariel

Sounds like a sexy design and perhaps a worthy alternative for those who have aspirations to collect the bits and pieces to build a M19.. GPA has now just released their own crossover for the M19, making it a one stop shop for all the bits except the horn etc..

I read somewhere that Bill at GPA says there is really no audible difference between the Alnico 416 and the Ferrite version, so maybe that could help shave the cost down a bit for some?

:cool:

I hate to disagree with Bill, but the Alnico 416B (Classic Series) sounds just amazing - going out on a limb here, but it's the best bass driver I've ever heard. The first thing you notice are the vivid tone colors of the instruments, and the in-the-room naturalness of the singers. I really wouldn't want to give that up to save $100 on the driver. Relative to the other costs of the speaker, including deluxe crossover parts, the AH425, and the compression driver, that's not really a major saving.

(Comment on magnet sound: if you can hear the difference between nickel-core, M6-core, and amorphous-core transformers, you most certainly will hear the difference between driver magnets. It's a very similar kind of sound. Alnico, to me, sounds a lot like nickel-core transformers, with lots of sparkle and immediacy to the sound.)

I've not heard a Model 19, but the hallmarks of this speaker are the delicacy and immediacy of the Alnico-magnet bass, the midrange effortlessness of the large-format compression driver, and the silky-smooth and unhornlike sound of the LeCleach' T=0.707 horn.

This may sound weird, but in terms of the presentation, the new speaker is kind of halfway between a wall-sized electrostat and a Klipschorn. Spatial impression is very deep, with performers in-the-room, but the performing acoustic going back about 30 feet or so. But the dominant impression is the vividness of tone colors, subtle dynamic shadings, and what seems like unlimited headroom (in reality, about 10 dB better than commercial audiophile speakers).

One thing I've learned from this project is that horns are far more sensitive to crossover adjustments than direct-radiator drivers. The low-slope (1st and 2nd-order) crossovers had more IM distortion than I could accept, even with the large-format compression driver. Twiddling with the corner shape was also quite audible as well - the more conventional Butterworth alignments had noticeable energy pile-ups near the crossover frequency, which impaired driver integration and led to a sense of spectral tilting. I gradually went to softer and softer corner-frequency shapes until the impression of energy pile-ups went away - basically, fiddling with the group delay until it fell below a perceptual threshold. Filters that are transitional between Bessel and Linkwitz-Riley 4th-order seem to work fine.
 
Last edited:
I hate to disagree with Bill, but the Alnico 416B (Classic Series) sounds just amazing - going out on a limb here, but it's the best bass driver I've ever heard. The first thing you notice are the vivid tone colors of the instruments, and the in-the-room naturalness of the singers. I really wouldn't want to give that up to save $100 on the driver.

That ebay price is quite attractive for a driver of that size with an Alnico magnet. :)

Do you have a 1 meter response for it, and T/S param.s?
 
The new loudspeaker is around 99~100 dB/metre/2.83Vrms, ...
The bass driver is an Altec/GPA 416B (Alnico magnet, 16 ohm version), the HF horn is an Azurahorn AH425, the large-format compression driver is either a Radian 745Neo (Neodymium magnet, 16 ohms) or Altec/GPA 288 (Alnico magnet, 16 ohms), and the passive crossovers ...
...
No inband equalization is required for either the woofer or HF horn system. The system is time-aligned with all drivers in-phase; the lip (outer edge) of the AH425, by lucky coincidence, is about 1/2" forward of the front face of the bass cabinet, with about 2" of vertical spacing between the top of the cabinet and lower edge of the AH425. The bass cabinet is about 4.5 cubic feet and uses a resistive vent.

No mention of the supertweeter here. Is it abandoned or not needed?
 
Part of the reason for a supertweeter is improved dispersion. The LeCleac'h T=0.707 horn with a 1.4" throat has audibly narrower dispersion above 5~7 kHz. It's not a laser-beam, but the off-center listener hears a moderately duller and less 3D pinpoint sound than the central listener. With a supertweeter coming in at 7~10 kHz, the sweet spot gets much wider - 3 listeners or more - while adding a touch of sparkle at the top.

The LeCleac'h horn has a completely different spatial presentation than a conical. There's a sort of searchlight on-off quality to the conical - if you're "in the beam", you're quite aware of the sound coming right at you, and when you move out of the beam, it seems to shut off suddenly. This is a subjective preference, but it's not a quality I like personally.

The LeCleac'h sounds more like a direct radiator, with soft edges, and no spotlight tendency at all. It's more like a circular electrostat, a traditional paper-cone tweeter, or a very smooth fullrange driver without a whizzer. The sound fills the room, rather than all coming right at you. Again, a subjective preference, but one I like.

There are several good candidates for supertweeter. Most of the horn supertweeters are inferior to the treble from the large compression driver, a testament to the quality of the aluminum diaphragm and good phase-plug design. I didn't try the exotic Goto and Ales supertweeters.

The super-high-efficiency ribbons like the top-of-the-line RAAL Lazy Ribbon are an excellent match, along with Heil AMT tweeters like the Beyma TPL-150 or Aurum Cantus Aero Striction, which also have high efficiency and high power handling. All of these match very well with the Radian 745Neo or Altec/GPA 288 Alnico, which have a much sweeter and relaxed treble than typical small-format compression drivers.

The trick is to select drivers with complementary dynamic characteristics. The Altec/GPA 416B Alnico calling card is dynamics and tone color. The Altec/GPA is different from modern prosound drivers; the voice coil is underhung, which means the field lines cutting through the coil are straight, instead of the curved lines that go through the outer portion of an overhung coil. The Altec/GPA cone and sticky-surround have very smooth out-of-band rolloff characteristics, almost like the much smaller Vifa 5.5" driver. This is not true for many modern prosound drivers, which need out-of-band notch filters or brickwall filters.

Compared to small-format drivers, the large-format compression driver has a much more relaxed presentation with an impression of unlimited headroom - the tradeoff is the top octave.

The supertweeter has to have enough subjective headroom to keep up with the large-format driver. Not only does this mean a high crossover frequency, but preferably a supertweeter designed for professional monitoring applications.
 
Last edited:
The super-high-efficiency ribbons like the top-of-the-line RAAL Lazy Ribbon are an excellent match, along with Heil AMT tweeters like the Beyma TPL-150 or Aurum Cantus Aero Striction, which also have high efficiency and high power handling. All of these match very well with the Radian 745Neo or Altec/GPA 288 Alnico

Lynn great to hear the project is at it's final stage and that such obvious solutions from the past with some today's knowledge work wonders.

Have you actually heard the above mentioned as supertweeters (the Aurum Cantus, Beyma...) or you are just judging on the technical data?

Also could you elaborate a little bit about the box and using the resistive port, how you came at it, it's performance etc..

Thanks getting back and continuing with this project
 
Administrator
Joined 2004
Paid Member
I understand the search for the super tweeter as I'm going thru the same thing with my Altec 288 drivers on the 1005 horns. Also coming in at about 7 KHz.

Been meaning to pull the trigger on a pair of Faital Pro HF-100 drivers with matching horns. Not too pricey at $50 each for horns and drivers - or $200 a pair all told. Funny that the horn costs as much as the driver. I've held one of the drivers in my hand (cute) and it seems well made, but not yet heard it.
 
10db of efficiency and headroom...

The super-high-efficiency ribbons like the top-of-the-line RAAL Lazy Ribbon are an excellent match

And there we were worring about saving $100 on the GPA bass driver....:D

But when you're looking at almost 10db of sensitivity over the Ariel and 10db or more in headroom while maintaining the quality of sound one might expect that the parts cost would grow similarly.
 
I understand the search for the super tweeter as I'm going thru the same thing with my Altec 288 drivers on the 1005 horns. Also coming in at about 7 KHz.

Been meaning to pull the trigger on a pair of Faital Pro HF-100 drivers with matching horns. Not too pricey at $50 each for horns and drivers - or $200 a pair all told. Funny that the horn costs as much as the driver. I've held one of the drivers in my hand (cute) and it seems well made, but not yet heard it.

If you go with either of Lynn's recommended compression drivers, and listen to them I am very confident you will be impressed with how good tney are
 
One of the primary goals for the large-format horn was the lowest possible diffraction, along with respecting the maximum-length limitations mentioned in the Newell and Holland book. As the project developed and Bjorn Kolbrek developed custom simulations of diaphragm loading vs frequency, another goal became constant (resistive) diaphragm loading over the working frequency range, without the severe peaks and dips of more traditional horns. This is when a T ratio of 0.707 emerged as one of the flattest diaphragm-loading profiles, with other profiles (including Tractrix) having somewhat rougher diaphragm-loading characteristics.

Following the Newell and Holland guidelines, uniform diaphragm loading, and the lowest possible diffraction, became the primary design goals, with other parameters falling by the wayside. That's the same set of priorities as N&H; they explicitly state that if other parameters are sought, such as constant directivity, then time response and acoustic flatness will suffer. The conventional techniques for creating constant directivity involve the intentional use of diffraction in the throat, which degrades time response and flatness.

Flatness (due to changes in diaphragm loading vs frequency) can be restored in the crossover, but once diffraction has happened (anywhere in the horn), the reflections cannot be cancelled by electrical means. There are FIR digital techniques for reflection cancellation, but the reflections unfortunately vary with emission angle, making the on-axis correction wrong for off-axis emissions.

My feeling is that diffraction is undesirable in any loudspeaker, since the resulting time errors are not correctable, even in principle. (What is right at one point in space is wrong everywhere else. The only exception would be a truly omnidirectional radiator with identical time response in all directions. Let me know when you find one.)

Following the N&H guidelines, directivity is allowed to simply be what it is. A side effect of time optimization is lack of sharp sidelobes (in the spatial domain), so the polar pattern has soft edges, like a direct radiator. In spatial terms, it sounds like a direct radiator, not like a horn, but in terms of dynamics, it sounds like a horn - a very powerful horn, since the large-format diaphragm has twice the area of a small-format driver, and the horn has 90~100% horn loading over the working bandwidth.

The debate over directivity is where I part company with my contemporaries. I put about 70~80% weighting of the direct sound, and look at the rest in terms of total power into a sphere, instead of what happens 30 degrees off-axis. That's because the first three reflections are off the floor, the back wall, and the side wall. Only the floor reflection is anywhere close to 30 degrees off-axis, while the rear and side-wall reflections are much further off-axis. The next ten to twenty reflections are also mostly very far off-axis.

When you consider the spectral response of the first three to twenty reflections, what matters is total power into a sphere, not the frontal radiation pattern. As for the first-arrival direct response, you can't be in several places at once. You only have two ears, a few inches apart. What matters is the direct sound arriving at each ear and the summation of the total power into the room, with a heavy weighting towards the first-arrival sound.

This is why I consider the absence of unwanted sidelobes important. This isn't a PA speaker, where audience coverage is the primary consideration. This is a speaker intended for high-quality listening in a domestic living room with domestic furnishing (no floor absorbers, no walls covered in damping or quadratic residue diffusors). If there are narrow sidelobes (characteristic of high-Q controlled-directivity designs), there's a good chance one of these might reflect from a sidewall, degrading stereo image quality. Better to have a soft-rolloff polar pattern, which is exactly what these horns do. The "edges" of the polar pattern are much softer than any other horn I've heard to date.

The caution with horn supertweeters is that using a conventional horn, particularly a constant-directivity type, results in a very different presentation than the AH425. That's what I heard when I auditioned commercial supertweeters with integral horn assemblies; the supertweeter said "look at me, I'm a horn!" while the AH425 said, "well, I'm a big electrostat, and I don't sound anything like you!"

I'd guess that a horn supertweeter is OK provided the horn is a small LeCleac'h with the same T ratio of 0.707, possibly as high as 0.8. Otherwise, it won't match spatially or tonally.

At this point you might think "How on Earth does a ribbon integrate with the AH425 and the large-format compression driver?" Very well, really, despite the huge difference in polar pattern. What's similar, though, is the time response. The AH425 has near-electrostat time response, with very fast decay characteristics, and the supertweeter needs to be as good or better - preferably much better.

If the time response is worse, I can say from experience it really draws attention to itself, and not in a good way. The ear is very sensitive to the time response imprint that a given driver imparts to the sound, and drivers with chaotic decay characteristics draw attention to themselves.

All three drivers (LF, HF, SHF) have very good decay characteristics (compared to the other drivers with similar efficiencies). They were selected for these characteristics, and share them with the drivers selected for the Ariel. Which is why the new system sounds like a (very) big Ariel.

So if you want to roll-your-own supertweeter, be my guest. Measure the decay of the time response and see how it compares to the competitors. The ones with the quickest, cleanest decay will probably be the best match for the AH425 and the Altec/GPA 416B in a resistive-vent cabinet.
 
Last edited:
Administrator
Joined 2004
Paid Member
Thanks Charlie. I thought Melon meant compression drivers for super tweeters, that's what I've had a hard time finding.

FWIW, I run a pair of Altec 288 (16 ohm) and also have a pair of older Radian large format drivers. Love 'em! Altec 416A are my woofers.