You brought me an interesting point. I've excluded hard dome materials because I rarely found ones sounding good to my taste: Scanspeak 98000 used in Revel, Accuton ceramic domes used in Marten, Artemis Eos and a DIY one, and Beryllium dome in a TAD, a Magico and a DIY one, and Usher DMD. All others I liked were textile domes. So, actually I want to make this new project modular so I can easily replace tweeter modules(and maybe midrange/midbass too). But as a beginning I want to go with familiar and safe textile dome.
Do you know of better textile dome thanks to its better dome shape?
In addition to the ATC mid, I have a somewhat strong preference with D2104 for now, but haven't heard of it yet, a better sounding one will definitely work as long as the price is not too high.
I actually want to go with modular design, separate enclosures for different drivers/ways. It doesn't need to be a full range as I have a 13" SVS sub. I don't think many people will call a speaker with 2 8" woofers a full range. ;=)
I think I've successfully used the DSP crossover by modifying the DSP F/W of a surround processor, and will use some professional gears for streaming processed audio data to multiple DACs, then all the processing is done by the powerful Apple M1 processor in a Mac mini.
It can be hard to give constructive advice when one is following enthusiasms rather than dry engineering (which is of course a perfectly reasonable thing to be doing with a hobby!). ATC stopped supplying their small soft domed midrange to other manufacturers and the DIY market a few years back so it will be difficult to obtain a pair and, depending on how obtained, ATC may not support them. The Volt drivers seem to be a small step down in performance and are a ludicrous price. If one really wants to use a small 3" soft dome the Neumann neo one seems to offer the same level of performance as the ATC it replaced but in a more usable package. Mind you I have not seen it used in DIY projects so it may be just as difficult to obtain as the ATC for DIYers. It's not something I have looked into. Anyone?
JH> I think I can get a pair of new SCM75-150 at an okay price. Definitely not in volume to do any meaningful business.
. I agree with you on Volt 3". I don't see much information on the 3" ones Neumann is using. Can it be the cheaper Scanspeak 3" midrange with Neumann's own waveguide design?It might work after a fashion but it won't be working well at the low frequency end compared to a more appropriately sized midrange. To work well it likely needs a larger waveguide and a higher crossover frequency to the woofers (or lower midrange + woofers). It also flaps a bit at high frequencies and so will likely benefit from a lower crossover frequency to a 1" tweeter in a waveguide. That is, more like how K&H used to use it rather than ATC.
JH> I was kind of curious about it as 380 ~ 3800 is a very wide range for a midrange driver. I once thought about using Volt's new 2" dome and a 4" ~ 7" cone(slitted cone or the illuminator one) as a mid-high + midbass design. But I kind of gave up that idea because my friends were against the idea of dividing the voice range over to 2 bands/ways, and the 2" dome didn't seem to behave as well as I hoped. Maybe you can tell me it is actually not a bad idea, and I might get excited to pursue the design.
A small tweeter gives wider dispersion not better dispersion. Better dispersion in a technical sense is more like smoothly controlling the variation in beam width with frequency to get as close as possible to a neutral timbre in use in a room. Of course better dispersion in an audiophile sense may, quite reasonably, mean whatever sounds good to me which may well be a rapid move to wide dispersion at high frequencies.
JH>. This is tricky to understand, but yes, I think I am equating 'wider' to 'better' here, because I find my current speakers' tweeter dispersion is quite narrow, and it at least doesn't measure well.
A soft 3" midrange driver can be made to work well but there are as good if not better and cheaper approaches which is why they have become rare these days. So if your objective is to design a speaker that uses an ATC/Volt 3" soft dome then yes that makes sense. If your objective is to design a high performance speaker then preselecting an ATC/Volt 3" soft dome driver (or any driver for that matter) makes less sense.
JH> It's actually more of a safe proven choice rather than some emotional attachment. But I think I am at least open to listen to alternative suggestion and the reasoning.
Attachments
JH> I think I can get a pair of new SCM75-150 at an okay price. Definitely not in volume to do any meaningful business.
It is a Neumann design to replace the ATC when it became unavailable. Don't know where it is manufactured. Anyone? It matches the performance of the ATC in use in terms of clean SPL and low distortion as can be seen by comparing measurements of current and previous K&H speakers with the ATC driver.
JH> I was kind of curious about it as 380 ~ 3800 is a very wide range for a midrange driver. I once thought about using Volt's new 2" dome and a 4" ~ 7" cone(slitted cone or the illuminator one) as a mid-high + midbass design. But I kind of gave up that idea because my friends were against the idea of dividing the voice range over to 2 bands/ways, and the 2" dome didn't seem to behave as well as I hoped. Maybe you can tell me it is actually not a bad idea, and I might get excited to pursue the design.
A passband of a decade is about right for a driver although a soft dome struggles more than most to achieve it due to a lack of damping from a surround and strong resonant motion within the passband causing mechanical and response issues. When the ATC was designed nearly 50 years ago tweeters were significantly less capable than they are today in terms of clean low frequency output leading to high crossover frequencies like 3.8kHz being required. It doesn't make the same sense today where a 4-5" midrange with a passband half-to-one octave lower can be expected to lead to a superior overall speaker performance.
A 400-4000 passband leaves 2 and a bit octaves for the tweeter to handle and 4 and a bit octaves for the woofer. It is rather unbalanced for a 3 way but fits a 4 way with distributed subs handling the lowest couple of octaves and 3 way mains. This is of course the approach taken by many that take sound quality seriously such as studios. Technical sound quality that is rather than audiophile sound quality (which can be just as valid depending on what one values).
A pair of 8" woofers in a modest sized cabinet and perhaps sealed, your 3" ATC in a deep waveguide covering perhaps 500-2.5kHz) , a 1" tweeter in a waveguide and 4 x 10-12" subs is likely to make a well balanced high performance system if well implemented.
A 2" rather than a 4-5" for a midrange in a 3 way is daft in terms of technical performance because it will distort at higher SPLs. However, in an audiophile sense the wide dispersion of a 2" on a flat baffle may be perceived as attractive if one only listens quietly and not at standard levels. PMC cross a non-waveguide 2" soft dome at 400 Hz in some of their 3 way speakers so it is easy enough to check the pros and cons.
A 3" dome midrange is far from the safe proven approach. It is expensive to get performing well and then has a limited passband. It can be made to perform well but in practice this seems to be only worthwhile for companies that have a long history of using it in their products and have some business incentives to continue such as customer expectations, lower development costs and the like. The safe proven approach is a 4-5" cone midrange used in almost all high performance 3 and 4 way speakers.
What's wrong with emotional attachment when pursuing a hobby? If you like the ATC soft dome then go for it. It's a driver with a number of positive attributes but also a few negatives (e.g. price, passband, size). Design to make best use of the positives and limit the impact of the negatives and you should have a good speaker.
Your attachments seem to be blank?
I agree with what was said before: active with DSP is the way to go, in that it's way easier to get the greatest sound. But how can this be accomplished with something like a Hypex FusionAmp when creating a 4-way? Are the two smallest drivers connected to one amp of the Hypex, and is there a passive XO for those two drivers?
Personally I think you should be able to get a fantastic sound with "just" a 3-way. If you choose drivers that have a wide useable range, you should be able to XO before any beaming or breakup. This might mean you'll probably lose some of the low/high end of the ensemble on paper (not the biggest woofer, not the smallest tweeter), but you can correct most of that with DSP.
Currently have two Hypex FA towers with a WO24P-8, 12MU8 and TW29RN-B, I think going 4-way would not add any clarity, just complexity.
Personally I think you should be able to get a fantastic sound with "just" a 3-way. If you choose drivers that have a wide useable range, you should be able to XO before any beaming or breakup. This might mean you'll probably lose some of the low/high end of the ensemble on paper (not the biggest woofer, not the smallest tweeter), but you can correct most of that with DSP.
Currently have two Hypex FA towers with a WO24P-8, 12MU8 and TW29RN-B, I think going 4-way would not add any clarity, just complexity.
I am in agreement with @andy19191 - A pair of high quality 8" woofers can cross at 500 - 600 Hz to a 3" dome mid such as the kind you are considering. If your woofer was a 15", then there would be a good argument for a 4-way, such as 15+8+3+1. But a pair of 8's should work very nicely with the dome mid.
Oooops, I was just told I missed out a sim on my PDF here is the missing "4th Sim" showing a single driver holding 103 dB @ 1 m with 8 watts which is just 4% (!!) of its AES rated continuous power. A pair in room will hold 106 dB at one meter with vanishing low distortion.
Crucially the cone movement is just +/_ 0.4 mm... Less than 14% of its Xmax!
This demonstrates the advantages of "a good big one Vs a good small one".
I believe minimising cone travel & input power to achieve a given SPL (106 dB in room is very loud!) are mission critical for realistic sound reproduction.
All forms of linear, non linear and thermal distortion (power compression) are minimised or eliminated and the most offensive of all distortions, Time Domain distortion is drastically reduced.
When choosing any driver I would urge you to carefully compare two key points:
(1) How much cone travel does it require to meet your target SPL in room over the frequency band its covering?
(2) How much power does it require to meet your target SPL in room over the frequency band its covering?
Remember point source SPL reduces by 6dB per doubling of distance so 106 dB at 1m = 100 dB at 2 m and just 94 dB at 4 m listening position.
over
When
Crucially the cone movement is just +/_ 0.4 mm... Less than 14% of its Xmax!
This demonstrates the advantages of "a good big one Vs a good small one".
I believe minimising cone travel & input power to achieve a given SPL (106 dB in room is very loud!) are mission critical for realistic sound reproduction.
All forms of linear, non linear and thermal distortion (power compression) are minimised or eliminated and the most offensive of all distortions, Time Domain distortion is drastically reduced.
When choosing any driver I would urge you to carefully compare two key points:
(1) How much cone travel does it require to meet your target SPL in room over the frequency band its covering?
(2) How much power does it require to meet your target SPL in room over the frequency band its covering?
Remember point source SPL reduces by 6dB per doubling of distance so 106 dB at 1m = 100 dB at 2 m and just 94 dB at 4 m listening position.
over
When
A superb commercial example of large Sd driver deployment (even twin AMT tweeters in a unique horizontal layout) is the new "Legacy VALOR" speaker system. At only $96K (!) it is one twenty'th the price of the usual suspects high end mega systems BS ie $1 million for pair of passive (!!) speakers and another $1 million for the electronics and cables.
" I’m an old school, analog, tube-loving, horn-speaker–espousing kind of guy. I'm not against, but I'm largely uninterested in, most new technologies. But when I sit in front of a streaming, active, class-D powered, DSP-enabled Legacy system, I never fail to relax. To feel engaged, alive. And charmed. I don't want to get up. I don't want to leave. I don’t know why Legacy founder and president Bill Dudleston’s systems sound so good and make me feel so good, but they always do." Ken Micallef - Axpona 2024 https://www.stereophile.com/content/legacy-audio
" I’m an old school, analog, tube-loving, horn-speaker–espousing kind of guy. I'm not against, but I'm largely uninterested in, most new technologies. But when I sit in front of a streaming, active, class-D powered, DSP-enabled Legacy system, I never fail to relax. To feel engaged, alive. And charmed. I don't want to get up. I don't want to leave. I don’t know why Legacy founder and president Bill Dudleston’s systems sound so good and make me feel so good, but they always do." Ken Micallef - Axpona 2024 https://www.stereophile.com/content/legacy-audio
JH> If the final performance is matching ATC's, I don't think the Scanspeak dome mid can be transformed to such only with a waveguide. Interesting to know more about it. I also found a driver vendor called Bliesma making big dome midranges, up to the size of 142mm! They also make 74mm ones with Be, paper(!) and silk diaphragms. I will see what people say about them.
JH> How about max SPLs of those small cone drivers? When I checked last time(not pro drivers), they seemed to struggle quite a bit in that aspect.
JH> I'd take this as an okay sign for the 4-way design with 2" dome?
JH> How about 4-way design with 2". I'd like to see your more detailed comment on that approach.
JH> What I meant was subjective and sonically proven in my experience. I am just a newbie who happened to have a subjective success only once, and not as knowledgeable as you in many physical aspects of mechanical stuffs. I wonder if my steep DSP crossover(100+dB/oct) is hiding the problems beyond my audibility. I listen to symphonies very loud(to my standard). At tutti, I think it approaches 100 ~ 105dB sporadically.
One interesting thing I just recalled about ATC: I have a mastering engineer friend who was once(at least) nominated to a Grammy award, and he strongly prefers active ATC studio monitors to anything else(except one) and always use them when budget permits. The exception is an old WE horn system actively crossover'ed and driven by tube amps. He has the most sensitive ears among people I've met, and his auditory capability is objectively proven. He never said an adverse things about ATC's sound. He also recommends ADAM and Elac for lower budget systems, so just not an addict to ATC.
I do like ATC mid, but am also curious enough to know about alternatives, especially since most people chiming in prefer something else. I would go with ATC if 3-way, but 4-way is also in my interest.
I didn't mean to attach anything and they were added erraneously. Then I can't edit the post, wth....
I thought I'd listen to music very loud, and may be I was wrong.
. I see that you use very linear regions of those "loud" drivers, but I had never been interested in very "pro-ish" drivers(ATC is kind of audiophile-ish, right?), with my space just a bit big for one in a home.One non-sonic issue of mine with your approach is WAF. I'd like to avoid making frontal area of the speaker(not the drivers, but similar) very large. Your approach seems to dictate what I want to avoid aesthetically, unless the woofer is mounted to side.
I like the Legacy's approach, but I have rarely heard satisfying sound from class-D amps. The only 2 exceptions were Lyngdorf/Steinway system using class-D amps with DSP crossover filters built-in, and some wacky GaAs-based class-D module housed in a glass tube(forgot the company name). I am using Monolith 9-ch HT amplifiers for the most part, and feel safe and better with old class-AB yet.
Last edited:
Yes, I was expecting this kind of answers. Maybe 2" dome is the only way to justify 4-way?
Thank you for chiming in on this. It's good that at least I have one part of this done right - DSP crossover. But, being not sure I understand your comment, I want to clarify that I am not using or planning Hypex or any class-D for my current/future speakers. They are all class-AB except my SVS sub, and the DSP crossover is done in a surround processor.
I also tend to agree 3-way is technically better with 2 8" woofers. That's why I asked this question to see if there is any yay answer.
I also am not a fan of any class D I have heard, but I would love to try the new GAN FET versions... Perfect for subs and mid bass with tubes on mid and top.I thought I'd listen to music very loud, and may be I was wrong.
One non-sonic issue of mine with your approach is WAF. I'd like to avoid making frontal area of the speaker(not the drivers, but similar) very large. Your approach seems to dictate what I want to avoid aesthetically, unless the woofer is mounted to side.
I like the Legacy's approach, but I have rarely heard satisfying sound from class-D amps. The only 2 exceptions were Lyngdorf/Steinway system using class-D amps with DSP crossover filters built-in, and some wacky GaAs-based class-D module housed in a glass tube(forgot the company name). I am using Monolith 9-ch HT amplifiers for the most part, and feel safe and better with old class-AB yet.
Re your WAF cabinet situation:
What is your approx max acceptable cabinet dimensions if its a floor stander and please indicate your preferred cabinet style and approx dimensions.
One can easily reduce the cabinet volume and play around with the shape and driver choice.
What about "On Wall" ?
ie Imagine a flat line array approx 1.5 m tall, 40 cm wide (anything from 30 cm to 50 cm is good) and approx 20 cm deep. They would sit on top of your subs and usually either side of a screen/TV if part of a home cinema.
Side mounted woofers can be good, but only if crossed over below 80 Hz, preferably 60Hz, but then the front to back depth of the cabinet increases.
It depends how you look at it. For example a subwoofer added to a 3 way equals a 4 way. I went large drivers 15" subs, 15" mid-bass, 10" midrange and a large format compression driver on a 100x100 CD waveguide. Crossed at 80Hz, 300Hz, 1.5K
Another example would be L250 14, 8, 5, 1 inch Dome as a classic passive 4 way.
Rob
Another example would be L250 14, 8, 5, 1 inch Dome as a classic passive 4 way.
Rob
Thank you for your further interest and advice.
I don't have a firm requirement or criteria, but I think my current one using 10" B2500.1 is a bit too wide. I want to make it taller, narrower, and probably deeper. 8" was chosen in a way for aesthetic reason. ;-). I also want to make this modular(separate box per driver or way), so sorry, things are a bit more complicated.
What do you mean by flat line array? Something like "column" type speakers or McIntosh's old ones? That's another interesting approach, but I want to explore it later. And, I don't have 2 subs, and don't want to place it/them in the front area below the front left/right, sorry due to the aesthetic reason, again.
If the speakers are made very deep, I guess it's practically on-wall...? That should be fine in WAF. If you meant a wide/tall/narrow ones by "on wall", I think my wife will say no. And thank you for the information that 80Hz or 60Hz is good and low enough for side woofers.
Oh, BTW, what about the tubular enclosures? I believe it's difficult to make, but stacking up the pipes with front-mounted driver is quite interesting to think about.
If you really want a 4-way, I am sure we (and you) can come up with a justification.
Normally, a 3-way plus stereo subwoofers can be considered a 4-way. So one option is two SVS subs (or equivalent), and then your planned 2x8 + 3 dome + tweeter. With stereo subs positioned reasonably close to the main speakers, the sub crossover could be raised up a bit, say 120 Hz. You would have a lot of dynamic headroom with 2x8's running from 120 up. 2x13 + 2x8 + 3 + tweeter, nominal crossovers at 120, 600, 3.5k... This sounds very plausible.
Now let's assume you want a 4-way not considering the SVS sub. Honestly it does not make sense to insert another driver between an 8" and a 3" dome. But it might make sense to insert a driver between an 8" and a 2" dome. The 2" could run from 1k to 4k, and the region below 1k could be split between the pair of 8's and some other driver or pair of drivers.
FYI ScanSpeak Revalator mid-woofers are fantastic and easier to integrate than the published FR shows, and also have very good dispersion and relatively good high frequency extension. For my center channel I saved money by not using Illuminator woofers so I could spend it on the tweeter. The Illuminator tweeter has amazing off-axis response, btw, but even the Discovery tweet sounds very very good.
I meant shallow, not narrow for on-wall, but I am not allowed to fix it....
If you really want a 4-way, I am sure we (and you) can come up with a justification.
JH>. Right. I meant reasonable technical justifications.
I wasn't including SVS when I said 4-way. I wasn't clear enough.
Yes. I agree with you on 3-way with 3" and 4-way with 2". My original question was kind of, "still.., just in case".
Among cone midbasses of size 4" ~ 7", which size would suit better between 2" dome and 2x 8" cones? Or maybe some specific driver(s) to recommend?
Hi Erik,
Thank you for your advice. Since you brought Scanspeak mid-woofers specifically, Which drivers do you recommend to bridge 2" dome and 8" woofers? I used to and still do like revelator slitted cones, but recently got impressed by the illuminator with some clover or flower-like shape on the cone surface.