To thereby keep the ME90 from having a wider dispersion than the midwoofers?
Subject: Center Speaker Recommendations for Main Horn Speakers
- Thread starter ajant
- Start date
These tradeoffs are driving me nuts!! But Earl Geddes recently mentioned that the smallest driver practical is best. https://www.diyaudio.com/community/threads/best-compression-drivers-today-2022.382609/post-8135700 But a 1" driver would rule out the M2 , ME90 and other waveguides but nominate which ones? And which are among the best performing 1" drivers? Or should I stick with the 3" or 4" drivers and the waveguides discussed?
Duke LeJeune
Manufacturer - Illusio Audio and AudioKinesis
Certainty is a rare commodity in this sort of venture.
If I was in your situation (lacking as much relevant personal experience as I'd like, and observing that the majority prefers one thing and one random manufacturer prefers something else), I'd probably play it safe and go with the majority.
The on-axis dippage I mentioned would be to compensate for the ME90's dispersion gradually widening south of about 1.5 kHz.
Because of this widening , there is a bit of a trade-off between good on-axis response and having a bit too much off-axis energy between 1.5 kHz and the 900 Hz crossover. My preference in situations like this is to "split the difference", and use a small on-axis dip to partially compensate for the excess off-axis energy.
I totally sympathize!
"No compromise" is a marketing department term. Loudspeaker system design is all about the juggling trade-offs.
Imo a good 1" throat driver has this main advantages over a good 1.4" throat driver: Given a suitable waveguide-style horn (something Earl invented so it's a given for him), a 1" throat driver can give you good coverage across a 90 degree pattern width well up into the top octave. There will be some pattern-narrowing in the top half of the top octave.
Imo the 1.4" throat approach trades off wide pattern coverage up high in exchange for being able to cross over lower. Like, you can get away with 900 Hz or maybe even 800 Hz with the ME-90 because 1.4" throat drivers have more output down there and are more rugged, but with a similar-sized 1" throat horn you probably would not want to cross over below 1 kHz at the lowest. On the other hand the ME-90's pattern narrows in the top octave, and I think it's because the 1.4" diameter inevitably starts beaming (in the absence of a diffraction slot).
Earl used a 1" throat driver, the B&C DE25 and later the DE250, on a 15" 90-degree waveguide and crossed over around 1 kHz to a 15" B&C woofer. I think your Altecs would be perfectly fine crossing over at 1 kHz, but my understanding is that you would prefer to have a lower crossover frequency based on the Altec's distortion measurements.
Earl later lowered his crossover frequency to 900 Hz, still with the 15" waveguide, and I THINK this is when he started using the DE550. The DE550 is among the few 1" throat compression drivers that have a 2" diaphragm. The others are the DE550TN (titanium rather than polymer diaphragm); the Kartesian Cmp50_vPA-S (titanium diaphragm); and the Faital HF1000 (polymer diaphragm). The final version of his Summa used the DE550 on a big 18" waveguide with an 800 Hz crossover frequency.
I am pursuing the development of a commercial product in this general category (big woofer; big constant-directivity waveguide with a fairly wide pattern; 1" throat compression driver). I'd rather not post the specifics in your thread, but you can shoot me a message if you'd like, as some of it might be relevant to your project.
Last edited:
When I allowed Troy to use my Altecs to design a crossover for a customer using his ES600 horn it was also at 800Hz. Note the listening tests, though discount the issues about the low bass as Gary Dahl designed the cabinets to only play down to ~ 70Hz to minimize distortion above his subwoofers, as I will be doing.
Custom DIY Plan Set for Altec 416-8B with ES-600 Biradial No.1978
In this blog post I would like to feature a custom order completed for a customer who had requested a DIY plan set for his Altec 416-8B 15" Woofers. He consulted with me on what would be the best 2-way configuration. I recommended the ES-600 Biradial No.1978 along with the newly offered SB...
josephcrowe.com
Any recommended waveguides which resemble Earl's 18" for use with the B&C DE550, or a similar driver with even better measurements to cross at ~ 800Hz? Earl obviously wouldn't recommend SEOS models.
Acoustic Horn Design – The Easy Way (Ath4)
It's not true - rectangular waveguides don't have to have problems on diagonal - I've just shown a counter-example above. And I measure an SEOS H & V and across the diagonal and the diagonal was the worst of the bunch. Are you saying that this is not true? That I screwed up the measurements?
Duke LeJeune
Manufacturer - Illusio Audio and AudioKinesis
When I allowed Troy to use my Altecs to design a crossover for a customer using his ES600 horn it was also at 800Hz. Note the listening tests, though discount the issues about the low bass as Gary Dahl designed the cabinets to only play down to ~ 70Hz to minimize distortion above his subwoofers, as I will be doing.
Custom DIY Plan Set for Altec 416-8B with ES-600 Biradial No.1978
In this blog post I would like to feature a custom order completed for a customer who had requested a DIY plan set for his Altec 416-8B 15" Woofers. He consulted with me on what would be the best 2-way configuration. I recommended the ES-600 Biradial No.1978 along with the newly offered SB...josephcrowe.com
That's great feedback on how your Altecs work with an 800 Hz crossover.
If somebody made and sold something like that, I think you and I would be among their customers.
It's a shame that JBL discontinued a driver that works that well with the M2, and which also has optional Be diaphragm compatibility. https://reconingspeakers.com/product/jbl-synthesis-4365-high-frequency-driver-476mg-440944-001/ https://www.diyaudio.com/community/...n-drivers-d2430-and-d2415.432445/post-8118758
JBL D2 Compression Drivers - D2430 and D2415
For clarification on the speculation on the D2430K drivers raised by Robh3606 There are two different part numbers 5032754X This version does not have a field replacement diaphragm part. If you do blow it up you obtain a replacement driver from JBL. This is the M 2 driver. It costs less than...
Btw, any experience with 18Sound drivers?
Not on the M2. Due to insane costs of Be products (presumably because of large volume consumption by the US military) over the last several years, I belieive that most DIYers use Radian Be drivers. Perhaps one can speculate over why Radian drivers are less expensive than other Be drivers. https://radianaudio.com/blogs/our-n...h-end-truextent-acoustic-beryllium-diaphragms But most driver experts like Arez cite 18Sound drivers as being especially well designed. Off hand
the only users I know of the https://www.eighteensound.it/en/products/hf-driver/1-5/8/ND4015BE is
https://www.diyaudio.com/community/members/drba.19834/ , whose uses it in his current speakers.
https://sphericalhorns.net/2023/07/15/drba-test-bench-eighteen-sound-nd4015be/ Presumably, at some point he may use it to test the waveguide he just finished prototyping.
IIRC, https://usspeaker.com/homepage.htm told me that 18Sound should have access to Be diaphgrams for their drivers that used them in another month or two. But 18Sound drivers with Be diaphgrams are very hard to find and/or sell for crazy amounts. I saw a pair of ND4015Be for ~ $7500. at a Euro store. But 18Sound drivers will always cost more than Radians.
Duke LeJeune
Manufacturer - Illusio Audio and AudioKinesis
The ND4015Be sounds like it a fantastic compression driver, based on that very thorough review. I hope it and/or the Radian make a comeback.
Of course they were, and will be, in a very different price category from the DE550 and its kind that we were talking about recently.
Interesting how much the lousy throat adaptor degraded the performance of the ND4015Be in the Voice Coil test. Thank you for providing those links.
Last edited:
You may want to check out my latest discussions with Earl, Art, Arez, Mayhem13 and Mabat. As for waveguides, says Earl:
I would not recommend any of those horns. To see great waveguide check out the ATH thread. Those people (Marcel) really know what they are doing. https://www.diyaudio.com/community/threads/best-compression-drivers-today-2022.382609/post-8140957
And so I went here and found what presumably are Marcel's three basic waveguides. https://at-horns.eu/
And here's Marcel's long extant thread, where you can also see my post. https://www.diyaudio.com/community/...he-easy-way-ath4.338806/page-851#post-7915064
Then these replies earlier today to Art, Mayhem13 and Arez, beginning here. https://www.diyaudio.com/community/threads/best-compression-drivers-today-2022.382609/post-8171246
Note that WAY back to square one idea about maybe (?) being perfectly able to use the 98db Altec 416's with something like this https://www.madisoundspeakerstore.c...nator-d3004/6640-00-1-tweeter-beryllium-dome/
http://www.troelsgravesen.dk/D3004-604010.htm
and an appropriately high performance midrange cone driver in a three way biamped system? This is certainly going pretty far afield from building horn speakers, but if Troy is unable or perhaps more likely unwilling to 3D print and robustly harden for drilling one of Marcel's waveguides, then maybe he would be wiling to build boxes for these dome and cone drivers and test the entire system with the DSP crossovers he likely has on hand. https://josephcrowe.com/pages/frequently-asked-questions-1#:~:text=combo.-,Do,possible
But how close to a constant directivity speaker system could be achieved-and one with the relatively high directivity index that Earl advocates to minimize secondary reflections-with this kind of a DSP crossed and biamped three way direct radiator system?
Please advise.
Thank you.
Last edited:
Duke LeJeune
Manufacturer - Illusio Audio and AudioKinesis
Yes, I am using Marcel Batik's math, with his permission, in my most recent waveguide designs. I'm not copying of his designs. But his math is the best I have come across for smoothly blending the mouth round-over to an Oblate Spheroid waveguide profile.
If you're going to make your own horns, or have them custom-made for you, then Marcel Batik's math is the best I know of.
Well, that would be a very different approach.
Imo using a dome tweeter and a cone midrange would be going in a different direction from what Earl advocates.
* * * *
I get the impression that you want a system which arguably "does it all", including (but not limited to): The right radation pattern width; good radiation pattern uniformity across as much of the spectrum as possible; low enough crossover point (500 Hz if I recall correctly) that your Altecs are being used where they perform best; 1" throat compression driver for good top-end performance; best possible compression driver diaphragm material; while avoiding the complexity and/or compromises of a three-way horn system.
I am not sure that all of these criteria can be met at the same time. In particular, I am skeptical about taking ANY 1" throat compression driver down to close to 500 Hz, and even if there is a 1" throat driver that would sound good crossed over that low, the waveguide would be huge. To a first approximation, you can assume that a 15" diameter waveguide is adequate for a 1 kHz crossover frequency and that as the wavelengths get longer, the required waveguide diameter grows proportionately. And if you want to use Marcel Batik's math, be aware that the bigger the waveguide, the more difficult it will be to fabricate.
On the other hand, in my opinion you would be quite safe crossing your Altecs over in the 1 kHz ballpark, which will give you better pattern-matching in the crossover region, and fabrication of a suitable waveguide becomes more feasible. I do not share your apprehension about the 3rd harmonic distortion of your Altecs north of 500 Hz, but please feel free to get other opinions on that topic (if you haven't already).
Last edited:
A center channel does not need to be a center speaker.
My recommendation is to find some nice easy to drive single driver full range bookshelf speakers....something like the ogy or the sibalious. Or a 8 inch ....but 2 speakers...even find a way they can switch from a center to a stereo with a small dedicated preamp and amp that can take the center or stereo input and switch when you want.
2 bookshelf speakers will give you a great ghost center if feed the mono signal and offer more versatility if you set it up right.
My recommendation is to find some nice easy to drive single driver full range bookshelf speakers....something like the ogy or the sibalious. Or a 8 inch ....but 2 speakers...even find a way they can switch from a center to a stereo with a small dedicated preamp and amp that can take the center or stereo input and switch when you want.
2 bookshelf speakers will give you a great ghost center if feed the mono signal and offer more versatility if you set it up right.
Nor would it even be my second choice as Earl's performance goals and design approaches are more sensible. And because the use of direct radiating HF and MF drivers with the 98db Altecs would likely require active crossovers, where despite their sonic advantages would have me living in fear of a high SPL accident, with the amplifiers connected directly to high efficiency drivers, especially in my Earl Geddes described small room.
Indeed, the ONLY reason I considered that alternate build was IF Troy Crowe was unable or unwilling to 3D print and finish the selected waveguide or if I could not find another trusted other service to do so.
Also, though crossing the Altec 416s at 500Hz would of course be ideal, Earl, Art and Troy have convinced me that 800Hz will not produce audible distortion, at least not at my average listening levels and distance.
But Arez replied with this. https://www.diyaudio.com/community/threads/best-compression-drivers-today-2022.382609/post-8172012
Ever hear of them? Evidently, quite popular among diyers. https://www.pcbway.com/why.html
Meanwhile, some encouraging feedback from Marcel!
Acoustic Horn Design – The Easy Way (Ath4)
mabat, prepare to be changed ;-) // After six years, it's about time for a student to surpass his teacher. :)
And here's the waveguide he tentatively recommends.
https://at-horns.eu/gen2.html
or https://at-horns.eu/twinray.html#twinray36-380
Last edited:
Duke LeJeune
Manufacturer - Illusio Audio and AudioKinesis
Sounds great!
I don't have any experience with PCB Way.
Meanwhile, some encouraging feedback from Marcel!
Acoustic Horn Design – The Easy Way (Ath4)
mabat, prepare to be changed ;-) // After six years, it's about time for a student to surpass his teacher. :)www.diyaudio.com
And here's the waveguide he tentatively recommends.
https://at-horns.eu/gen2.html
or https://at-horns.eu/twinray.html#twinray36-380
Marcel of course knows his own waveguides better than I do.
My personal inclination would be towards his most constant-directivity waveguides: https://www.at-horns.eu/A460D.html
Imo maximizing constant directivity can call for a bit different approach to system EQ, with a gentle reduction at high frequiencies being arguably desirable. Quoting Earl Geddes:
"Only once did I ever deviate from the "ideal" and that was to lower the HFs a few dB because I came to realize that CD [constant directivity] puts way more HF energy into the room than normal and that makes the speakers sound bright."
Last edited:
Simpleton question: What frequency range do y'all think is an acceptable 'minimum' to accomplish the task of a center channel speaker?
If I was forced to guess, I'd say 75-80Hz to 8-10K, primarily for enhanced vocal perception?
If I was forced to guess, I'd say 75-80Hz to 8-10K, primarily for enhanced vocal perception?
Sounds about right, might want a little lower depending on the overlap with the sub/subs and your crossover points and slopes available in your receiver or processor. I would also go out beyond 10k especially if the mains had extended upper F/R. You want seamless L/C/R pans, too much of a difference could be audible.
Rob
Yes, IIRC, Earl always sought maximal constant directivity, with also with a relatively high DI, yes? If so, does that, in fact, equate to the downward frequency response which you advocate for such a directivity pattern?
Also, Marcel mentioned here https://at-horns.eu/gen2.html the A460G2, which presumably is the revised version of the A460D. https://cults3d.com/en/3d-model/various/ath-a460g2
A460G2 is the latest iteration, maintaining virtually the same overall performace of the large A520G2 down to around 600 Hz,but in a noticeably smaller format (using a slightly different shape of mouth termination). It can be recommended especially for 1" drivers in general,where a bigger device offers typically only a negligible advantage.E520G2 is a segmented version[3] of the A520G2. Performance wise, it is on par with the axisymmetric version, with possible practical advantages.
Would you therefore instead recommend the A460G2?
Duke LeJeune
Manufacturer - Illusio Audio and AudioKinesis
The downward frequency response is not inherent to constant directivity with a relatively high directivity index. It has to be either an aspect of the speaker design or achieved via equalization.
I don't think there is a "wrong choice" among Marcel's waveguides. My understanding is that his Gen 2 waveguides have the most well-rounded combination of attributes, and the most adaptability.
Personally I would prioritize the constant-directivity characteristics of the "True-CD" waveguides, but that just happens to be my opinion. I think you will have a superb system either way.
Duke LeJeune
Manufacturer - Illusio Audio and AudioKinesis
Imo you want to end up with a frequency response that looks more like a top-end "tilt" than a top-end "shelf". But exactly what filters are used to get there depend on the specifics.
Also, ask your crossover designer to take into account your intended listening axis, if it's not directly on-axis. Earl designs for a listening axis that is 20 degrees off to one side, so 20 degrees off-axis. Since "constant directivity" is seldom PERFECTLY constant, there is usually a little bit of a downward-tilt to the top end at 20 degrees off-axis, relative to directly on-axis.
Similar threads
 | Steve Williams Site Founder | Site Owner | Administrator |  | Ron Resnick Site Owner | Administrator |  | Julian (The Fixer) Website Build | Marketing Managersing |