Why don't people build more coaxial systems?

^ well, it's not really an issue if the coax drivers are not asked to go very deep. For small speakers I have a pair of active woofers which do very well, and since bass is mostly non directional, so long as the level is smooth over the frequency range ( no bumps or dips in the mid to deep bass crossover region ) then the imaging of the small or coax drivers works very well. Also, without having to deal with the deep bass, they become a lot more dynamic.

In this, the AE1 are a rare bird in that you really want to run them full range... deep bass will just move the speakers about but you won't hear... however they are designed to handle the power. I haven't driven the Unifi 2 that hard but I suspect they can easily handle an Aleph 5. I know they can handle the power of a good 90 wpc SS class A/B, running full range, although the bass does get a little muddy.

The speakers I'm so curious about though, the MoFi SourcePoint speakers.. they are supposed to go deep and they are Andrew Jones babies too.

At 91db/w they seem like they'd be Class A friendly... but I worry they recommend a minimum of 30W... why? And not that I care that much for The Julian Hirsh approach but the Stereophile graphs show a pretty smooth frequence response at the crossover. Red is the SourcePoint 10, blue is the Mission 770 speaker.

View attachment 1294366

https://www.stereophile.com/content/mofi-electronics-sourcepoint-10-loudspeaker-measurements
By my understanding one of the main problems with full ranges and by extension the coaxials is that the cone is also a waveguide for the highs. The Doppler effect from the cone movements distorts all high frequencies, so we must limit the excursion at all costs. If the driver is capable of sub-100 Hz frequencies it's no help, it's better be high passed. 100 Hz is 3 m, it's extremely omni, it may be handled by a sub located anywhere.
 
  • Like
Reactions: GM and Juhazi
in directivity it’ll be the other way around
Agree. Almost deserves a somewhat romantic or Euphoric statement.

With right combination of filter orders asymmetric or symmetrical
Coaxial Can be considered almost dream like Vertical for anybody familiar with common center to center challenges of separates.

The challenge is finding a smaller tweeter that performs well inside a coaxial system.
Been exploring options. Think a 4" driver and 3/4 tweet offers best tradeoffs.

Then separate large woofer for low frequency performance. Which pose little to no phase issue at lower crossover points.
 
  • Like
Reactions: GM and otto88
By my understanding one of the main problems with full ranges and by extension the coaxials is that the cone is also a waveguide for the highs. The Doppler effect from the cone movements distorts all high frequencies, so we must limit the excursion at all costs. If the driver is capable of sub-100 Hz frequencies it's no help, it's better be high passed. 100 Hz is 3 m, it's extremely omni, it may be handled by a sub located anywhere.

True in principle. but you’re assuming the effect is very bad and very obvious

Like many things, the execution can reduce the inherent limitations. (Refer rear engine + a lot of power, and Porsche 911s).
I can’t recall people saying Kef’s or MoFi’s have poor treble

Here’s data - A Fly in the ointment - on a well-designed coax.
https://hificompass.com/en/reviews/satori-mt19cp-8-75-coaxial-speaker-review#13

While in the specific tests he did conclude “the parasitic frequency modulation caused by the Doppler effect is quite significant and is stronger the greater the amplitude of the midwoofer cone oscillations and the higher the frequency of the tweeter signal” …

Overall “The upper bass and midrange are excellent … … …… and

“The highs are not far behind the low and mid frequencies in terms of quality” so it can’t be too bad

you said we must limit the excursion “at all costs”. Absolutely, it’s desirable.

But he’s the most thorough driver reviewer around, and in his conclusions did not mention the Doppler effect. So its effect isn’t nearly as evident as his data suggests.

Me, I’d always want to use a dedicated woofer with a mid-woofer, whether it’s co-ax or not. (Except if it were a Purifi)
 
4" driver with 300 Hz or higher crossover hardly moves at listening levels.

A12pw-MTM-comp.jpg


dave
 
  • Like
Reactions: GM and WhiteDragon
Erin (bikinpunk here) tested KEF coaxials, tweeter response with woofer cone steadily at +/-3mm tested with KEF Q100 driver
Scroll down...
https://web.archive.org/web/2018030...ker-driver-tests/other-speakeer-driver-tests/


1736231484816.png


Another (closed) thread about coaxials https://www.diyaudio.com/community/threads/coax-woofer-cone-movement-impact-on-hf-response.275806/ with comment by Anrew Jones. Doppler effect is not a problem, that is how sound pressure variations (waves) propagate in air, despite of how they are excited! And that is how they appear in waveform display in oscilloscope or eg. Audacity if you zoom enough


1736232540918.png
 
  • Like
Reactions: otto88
Sound is pulsating air pressure variations and usually we can visualize it as waveform display in oscilloscope or eg. Audacity if you zoom enough. Different frequencies always ride on each other and modulate, this applies to all sounds, also from "natural" sources. The question is if a loudspeaker system makes any changes to the original stimulus.

Doppler distortion is modulation of pitch (frequency) https://en.wikipedia.org/wiki/Doppler_effect
Intermodulation distortion is added harmonic distortion, we can see them in multi(sine) tone tests

Audibility question with loudspeakers
https://pubs.aip.org/asa/jasa/artic...y-of-Doppler-distortion-in?redirectedFrom=PDF

And I agree that we should use a woofer (or several) with coaxial mid-tweeter! Been there too, done that.

https://www.audiosciencereview.com/forum/index.php?threads/kef-r11-meta-tower-speaker-review.53282/
"R11 Meta roared to action, delivering optimal and super enjoyable response on every reference track I threw at it. Science and excellent engineering works!"
 
  • Like
Reactions: GM and otto88
For the OP’s original question……For starters, it’s beyond a niche market…….drivers for DIYers alone are barely sustainable.

If I were developing a coax for market, since the drivers are already chosen to work together, I’d task a design for an optimal crossover and offer that as well making it less of a challenge for the end user.

Now the brutal truth?…..there’s perception and then there’s measurements. My own perception is that COAX drivers are lacking in clarity of upper midrange and low treble at volume compared to their counterparts. At normal volumes in the near to near mid field…..excellent……but they can’t compare beyond. Doppler distortion I suspect is the issue.
 
  • Like
Reactions: GM
Has anyone brought up car audio "coaxial"? DSP time-delay tweeter to be coherent with cone. Or do my reflector thing.

There are also many brands from Taiwan and China that sell home-hifi coax drivers or speakers of all sizes at different price points.
 
Last edited:
Excellent summary Juhazi, nails it. … and encourages me further

I guess you’ve used a few of the better coaxials, and know quite a few more … do you have some you'd recommend or particularly liked.
In particular for very clean sound. And sometimes at “highish” volume with “ample” headroom and low compression.
Quantitatively say target maximum output at max long-term power of say 110 dB/ 1 m or more. I’d think that would be a good indicator of ample headroom for the occasional favourite album listened with “clean” peaks of 105 dB at around 3- 3 1/2 m

I ask, as recently for the first time I developed a strong urge to do a coax.
Where very clean sound/ low distortion is very much the dominant criteria.

Other criteria … pretty open.
preferably dipole. but as it’d be with one or two woofers it wouldn’t need higher Xmax to help out at the low end.
Efficiency not important by itself - I’d use whatever good amp of adequate power to reach the SPL goal for whatever is the driver’s sensitivity
Frequency of xo to the woofer - from 100-300 Hz.
Size anywhere from 6 to 12 inches.
If the FR isn’t quite the smoothest, could be mostly dealt with by the DSP used for the xo.

As you’d know, at least half a dozen very good manufacturers have one or more drivers meeting most of the criteria.
Mostly pro drivers.

What comes to mind?
 
For loud/PA use check pro drivers 10-15" (some B&C used in this design by kimmosto, VituixCAD man). I can't recommend dipoles for high spl.
https://aaltospeakers.fi/en/product/taipuu-speakers-3-tie-dsp-aktiivikaiutinpari/

My latest diy project used a 12" coaxial and it was very difficult... the horn made problems.
Pics here https://photos.app.goo.gl/hGFkqajgoVwWrzL59
1736257632078.png
 
given the combination of distance and desired SPL, otbe (other things being equal), a 12 makes most sense

I’ve only read part of your thread, but from your experience likely to play safe and exclude horns

From memory, 12 inch pro coaxials from good brands … half a dozen or a few more. I should do a shortlist.


otbe, one with a low Fs/ Qms. Often said to be the best of the older 15 inches, I just remembered that the Altec 416’s Fs/ Qms of only 3.5 indicates very low loss > less inertia/ better control > better micro details (and performance at the lower end of the SPL scale).
Unless I find some comparative experience with “other things being equal” inc the parameters can’t be separated as likely have a better SQ.
 
For the OP’s original question……For starters, it’s beyond a niche market…….drivers for DIYers alone are barely sustainable.

If I were developing a coax for market, since the drivers are already chosen to work together, I’d task a design for an optimal crossover and offer that as well making it less of a challenge for the end user.

Now the brutal truth?…..there’s perception and then there’s measurements. My own perception is that COAX drivers are lacking in clarity of upper midrange and low treble at volume compared to their counterparts. At normal volumes in the near to near mid field…..excellent……but they can’t compare beyond. Doppler distortion I suspect is the issue.
But is that with a coax like the one in LS50 or the one in R3? Because one is a midwoofer, the other a dedicated midrange. And I can't really fault the R3 midrange version.