Maybe I should add that there are good coaxials out there but no one can buy them for DIY purposes?
There are plenty of coaxials on offer:
https://www.madisoundspeakerstore.c...rtQk7za6GwJp3xbyvYOp8MEkvukwo6aEDsg6AkhrpHtUk
https://usspeaker.com/coaxial speakers-1.htm
https://www.madisoundspeakerstore.c...rtQk7za6GwJp3xbyvYOp8MEkvukwo6aEDsg6AkhrpHtUk
https://usspeaker.com/coaxial speakers-1.htm
Last edited:
I'd say a few of them are available and quite nice.
I have hands on experience with TB W6-2313, which a very nice driver and a beast at imaging. Quite impressive.
I did not try the SBA SATORI MT19CP-8 myself yet, but I am quite sure it is a wonderful driver as well. SBA doesn't have many duds in their lineup. It's on my very short list of drivers to try.
The new W6 and W8 TB line of coax is very nice. Another set has been used in X's TL design and has the owner liking them very much.
SBA, I'm sure, is just as good. I should be able to prove it soon.
So, yes, there are good coax out there.
The TB drivers are exceptional as far as coaxials go. I would love to hear a comparison of their coaxials and their full ranges. That would be interesting. I wish I could come up with a couple grand to do that experiment.
Here is the typical Coaxial FR from the Satori MT19CP-8 Coaxial with Papyrus Cone. This kind of data scares me away from attempting to build a coax system. The frequency response above 2K is horrible.
I think there is a lot going on there that an average hobby hifi guy like me can't fix. The coaxial TB graphs are much better.
Here is a somewhat over-smoothed but much flatter looking frequency response from the W6-1916 6.5" Paper Full Range. Interesting to note that the TB coaxials also 'measure' very smooth. The TB frequency smoother is well known. This speaker is something the average guy can plop in a box and have reasonably good sound straight away.
This doesn't take into account the difficult to measure aspects of "cymbals appearing in the speaker cone" and "sound stage" that people can hear but modern machines for some bizarre reason can't measure. I think illusion and psychology are stronger factors than measured data. That's just my observation, don't go losing your mind and tell me I'm ignorant and don't understand physics.
My experience with several SEAS coaxials I admit is dated, the new designs from different manufacturers show promise, especially the TB offerings that seem to be well designed. They appear to be a system, not a tweeter stuck on a woofer. They appear to resemble other 'high-end' designs.
One last point before I hop on my tractor, there are planar loudspeakers that produce wide frequency response and are quite remarkable when you think about it. There are some rather large geometry differences, but I don't see why a cone can't be a good full range too.
Here is the typical Coaxial FR from the Satori MT19CP-8 Coaxial with Papyrus Cone. This kind of data scares me away from attempting to build a coax system. The frequency response above 2K is horrible.
I think there is a lot going on there that an average hobby hifi guy like me can't fix. The coaxial TB graphs are much better.
Here is a somewhat over-smoothed but much flatter looking frequency response from the W6-1916 6.5" Paper Full Range. Interesting to note that the TB coaxials also 'measure' very smooth. The TB frequency smoother is well known. This speaker is something the average guy can plop in a box and have reasonably good sound straight away.
This doesn't take into account the difficult to measure aspects of "cymbals appearing in the speaker cone" and "sound stage" that people can hear but modern machines for some bizarre reason can't measure. I think illusion and psychology are stronger factors than measured data. That's just my observation, don't go losing your mind and tell me I'm ignorant and don't understand physics.
My experience with several SEAS coaxials I admit is dated, the new designs from different manufacturers show promise, especially the TB offerings that seem to be well designed. They appear to be a system, not a tweeter stuck on a woofer. They appear to resemble other 'high-end' designs.
One last point before I hop on my tractor, there are planar loudspeakers that produce wide frequency response and are quite remarkable when you think about it. There are some rather large geometry differences, but I don't see why a cone can't be a good full range too.
They are indeed.
I think one of the reason they are smoother is the much flatter cone profile, compared to the deeper typical woofer cones, like on the Satori. Would probably be a reason for the erratic behavior of the tweeter inside the cone for the Satori.
One more reason is that TB is well known for doctoring the published FR. 🙂
Other TB I have tried are a W4-1052SD in the TABAQ which sounded nice with one notch, a w5-1611saf, that dropped like a brick past 5kHz, so definitively needed a tweet, and the woofer performance was not that stellar either, and the quite nice W8-1772.
That W8-1772 is a very nice one, with treble reaching far higher with a lot less beaming than most 8". The very thin paper cone reacts very nicely, but it desperately needs at least two notches to sound great. In fact, I haven't come up with any TB driver (or any driver from any manufacturer) that didn't need some kind of correction.
Any description that include quotes like "this one has a lot of air", or "details are prominent in this driver" will steer me far away or I know I will need to tame those things to enjoy them.
So, yes, I need to do some type of compensation. I just can't take out a driver out of its box, put it on a baffle and enjoy right away. I guess I'm picky! 🙂
Dave, can you link to some threads about DDR? Like data prooving it actually exists outside your imagination? How can it be measured? Since different fullrange drivers have different DDR, like MA good, TC9 no good, according to you, there should be some data supporting the claims.
In the near field, full range drivers with a benign break up mode past the listeners effective range of hearing and with suitable bass support can sound fantastic…………but outside of this use case, a well designed 2 way with a tweeter will outperform them at every turn.
Wanna experience the best Coax has to offer when you don’t have any measuring equipment or ability?…..grab a set of KEF Q150 second hand…..add some bracing to the cabinet and laminate the outside with a sandwich of premium plywood and constrained layer damping. Use the included port plugs and cross them to a suitable subwoofer at 80hz…….done.
Wanna experience the best Coax has to offer when you don’t have any measuring equipment or ability?…..grab a set of KEF Q150 second hand…..add some bracing to the cabinet and laminate the outside with a sandwich of premium plywood and constrained layer damping. Use the included port plugs and cross them to a suitable subwoofer at 80hz…….done.
@mayhem13 That's what I meant that they aren't available for the DIY crowd. You have to buy an existing speaker system to DIY your own.
I checked out the links by @Sonce and it looks like Faital has some great offerings in their pro line-up. The 10HX230 and 12HX500 look pretty good. https://usspeaker.com/faital pro 12hx500-1.htm
I checked out the links by @Sonce and it looks like Faital has some great offerings in their pro line-up. The 10HX230 and 12HX500 look pretty good. https://usspeaker.com/faital pro 12hx500-1.htm
It depends on how many moving parts you can manage skillfully in your design. I’ve used both FR and coaxials in numerous designs I’ve posted about in this forum.
If you want a simple concept and you are OK with being stuck with whatever attributes the FR driver itself has, then a full range is fine. Most people who build full range designs are looking for a very Zen-like minimalist setup.
But if you are a skilled designer, and you know how to use crossovers, DSP, active amplification and all the rest, then hands-down I think coaxial is superior.
If you want a simple concept and you are OK with being stuck with whatever attributes the FR driver itself has, then a full range is fine. Most people who build full range designs are looking for a very Zen-like minimalist setup.
But if you are a skilled designer, and you know how to use crossovers, DSP, active amplification and all the rest, then hands-down I think coaxial is superior.
Who can say with confidence which coaxial models are properly tweeter/midwoof time-aligned? Please name them to save me and others from making expensive mistakes.
If not time-aligned there's no way (short of DSP delay) to make it coherent and image as deep a soundstage as a good fullrange driver. If one were to use DSP delay, might as well bridge-mount favorite/SOTA 1.5" car tweeter backed with wool.
I tested Tannoy Precision 8 and KEF LS50 Meta 1793 drivers for coax time-alignment (not with usual precision due to size), by playing tweeter from one unit and midwoof from the matching unit, at multiple frequencies etc. The KEF passed, and I made a coherent 3-way LX Metals (in Gallery) but not without significant effort -- due to the waveguide and cone-horn mutually splashing sound all-over. And VHF still very much beaming.
MEH is a diy near-point-source-coaxial with greater flexibility. Wide VHF sweetspot but harder than offset 2-way-with-vertical-combing to make (passively) coherent.
If not time-aligned there's no way (short of DSP delay) to make it coherent and image as deep a soundstage as a good fullrange driver. If one were to use DSP delay, might as well bridge-mount favorite/SOTA 1.5" car tweeter backed with wool.
I tested Tannoy Precision 8 and KEF LS50 Meta 1793 drivers for coax time-alignment (not with usual precision due to size), by playing tweeter from one unit and midwoof from the matching unit, at multiple frequencies etc. The KEF passed, and I made a coherent 3-way LX Metals (in Gallery) but not without significant effort -- due to the waveguide and cone-horn mutually splashing sound all-over. And VHF still very much beaming.
MEH is a diy near-point-source-coaxial with greater flexibility. Wide VHF sweetspot but harder than offset 2-way-with-vertical-combing to make (passively) coherent.
Last edited:
its so much easier with a fullrange to get a linear phase than with two way concepts.
Indirect tweeters easily correct for off axis response.
Some easy alu foil covering of standard paper cone catapults fullrange drivers into another sonic league.
With current drive f response faults of fullranges somehow become less important.
But I have to admit that I heard very good loudspeakers from every category be it multi ways or horns.
Indirect tweeters easily correct for off axis response.
Some easy alu foil covering of standard paper cone catapults fullrange drivers into another sonic league.
With current drive f response faults of fullranges somehow become less important.
But I have to admit that I heard very good loudspeakers from every category be it multi ways or horns.
That is 2-way loudspeaker, by definition. Also screws up the phase.
Another reason I prefer a coax and active filtering is the issue of low frequency cut off. When you put a driver in a rear loaded horn, it greatly extends the dynamic range, but the benefits rapidly reverse below the cutoff frequency of the horn; and if you don’t have a appropriate filter, then the driver gets overloaded with a bunch of unnecessary low frequency garbage. All your XMax gets thrown away.
This issue is greatly exacerbated with a full range driver, where excessive excursion gets turned into intermodulation distortion in the treble.
Yes, even a zen purist full range system should have an appropriate high pass filter that is exactly tuned to the bass reflex, transmission line, or horn frequency cutoff. Very few FR designs have that.
Unfiltered designs are flapping in the breeze and nobody notices. It’s FAR worse in analog systems with turntable rumble. An appropriate filter extends dynamic range by 5-10dB.
I detail DSP<>full range filter design in an article called “the DSP assisted reflex“ which I published in AudioXexpress magazine
https://audioxpress.com/files/attachment/2721
This article shows an example of a system with a 3 inch Fountek driver in a 1 liter (!) passive radiator enclosure that reaches down to 55 Hz and plays a remarkably loud.
Many people will be quite surprised at how loud and authoritative a small FR or coax driver can sound in a reflex, transmission line or horn, when the high pass filter that protects the driver is precisely tuned to the system.
This issue is greatly exacerbated with a full range driver, where excessive excursion gets turned into intermodulation distortion in the treble.
Yes, even a zen purist full range system should have an appropriate high pass filter that is exactly tuned to the bass reflex, transmission line, or horn frequency cutoff. Very few FR designs have that.
Unfiltered designs are flapping in the breeze and nobody notices. It’s FAR worse in analog systems with turntable rumble. An appropriate filter extends dynamic range by 5-10dB.
I detail DSP<>full range filter design in an article called “the DSP assisted reflex“ which I published in AudioXexpress magazine
https://audioxpress.com/files/attachment/2721
This article shows an example of a system with a 3 inch Fountek driver in a 1 liter (!) passive radiator enclosure that reaches down to 55 Hz and plays a remarkably loud.
Many people will be quite surprised at how loud and authoritative a small FR or coax driver can sound in a reflex, transmission line or horn, when the high pass filter that protects the driver is precisely tuned to the system.
Last edited:
Here is my experiment with putting coaxial tweeter on fullrange and justification for doing so. Mainly for fun.
http://edsaudiopages.blogspot.com/2020/02/full-range-with-tweeter.html
http://edsaudiopages.blogspot.com/2020/02/full-range-with-tweeter.html
This is very true but it's also very much recording dependent. My older jazz CD's don't have that much deep bass in them and I don't usually crank them up on my full range speakers (an oxymoron). They sound really good from a 4" FR in a TL. When I want to play Saint Saens 3rd then I have to go with a system with several subwoofers and lots of power. Also, I only have digital sources most of them have been edited/re-mastered and the rumble of the old analog noise is largely removed. In many cases there is nothing there to filter. In some "classic" recordings there is traffic noise and air conditioner noise. So, it depends.
I don't think most coaxials, if used full range with a 3-5K crossover, will remove all of the IMD. It's reduced, but many of us don't hear much treble so we hear the IMD in the upper midrange which is near the crossover. I think a lower crossover is better but you can't get that with many coaxials on the market. Adding a woofer greatly reduces IMD in any speaker system (two way or full range) I'm a strong proponent of the WAW or large two way loudspeaker.
I have measured my hearing and my old tactical comm ears don't hear much above 12,000hz so I don't see the purpose of extended high frequency drivers. Super tweeters are pointless. I don't think I've ever heard "air". Also, I don't hear anything below 32hz either, but I do like the physical sensory rumble of big drums and huge pipe organs. So for me a full range with a subwoofer works better and is much less expensive.
[In my opinion records are the absolute worst analog source ever, I don't understand the fascination with turntables. Audiophiles complained mercilessly about how awful records were for years before CD's came out. Now we have HiRes digital recordings and people want them pressed into a record. It's insane. Anyway that's a topic for a different thread.]
Except you are adding in a splash of time-delayed (and lower-level) HF bouncing off the ceiling or wall. It is quite diiferent then the phase issues with a front-faced physically (displaced still) XOed tweeter.
Sometimes the efect works well — a semi-omni. Similar, but the castle microTowers work really well in some situations (front & upfacing FR).
But with omnis and cousins, the ones that belive the measures (of the smoothnness of the surface) outweigh their ears — they are telling, but not the whole story — will freak.
dave
The Linn DMS Isobariks from the 80s were like this. Midrange and tweeter on front and top. I'm sure the measurements were disquieting, and the imaging was lousy... but the speakers did sound great.
@Sonce
you do not screw up the phase if you have the indirect tweeter doing nothing below 5khz, 18db filters can help to achieve this.
Just read here about it:
https://en.wikipedia.org/wiki/Precedence_effect
The tweeter should have directivity properties, and if you choose beaming to the sides, up or behind the box in order not to interfere with the wave from the front it is not harming the direct wave.
The ear is not very good in phase perception above 5khz.
Its true, with an indirect tweeter the fullrange becomes by definition a two way.
However indirect tweeters do not need to be time aligned like in a two way where the tweeter beams with the main driver together to the listener.
you do not screw up the phase if you have the indirect tweeter doing nothing below 5khz, 18db filters can help to achieve this.
Just read here about it:
https://en.wikipedia.org/wiki/Precedence_effect
The tweeter should have directivity properties, and if you choose beaming to the sides, up or behind the box in order not to interfere with the wave from the front it is not harming the direct wave.
The ear is not very good in phase perception above 5khz.
Its true, with an indirect tweeter the fullrange becomes by definition a two way.
However indirect tweeters do not need to be time aligned like in a two way where the tweeter beams with the main driver together to the listener.
Time coherence is most important in the frequency region between 200 hz and 5khz.
So the "fullrange" concepts with a main driver going up to 4 to 5 khz and a tweeter doing the rest above works pretty well concerning imaging.
There are some fullranges which have no beaming at all and give a wide dispersion.
Like BMR drivers. I built a fullrange 8 inch giving wide dispersion and smooth response off axis.
So the "fullrange" concepts with a main driver going up to 4 to 5 khz and a tweeter doing the rest above works pretty well concerning imaging.
There are some fullranges which have no beaming at all and give a wide dispersion.
Like BMR drivers. I built a fullrange 8 inch giving wide dispersion and smooth response off axis.