• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Crazy PA Coaxials for Hifi

tmuikku

Senior Member
Joined
May 27, 2022
Messages
301
Likes
336
Look for VituixCAD measurement manual. If you perform messurements and convert impulse responses into frequency responses keeping the windowing same between all files, like the manual suggests, the timing info is preserved in the data and you dont need to fiddle with acoustic offsets and related issues, they are already included in the data.

Perhaps you knew this, still wanted to make the remark for future readers.

And yeah, DSP has far more positives for sound quality than negatives, even counting in cost of extra amplification and cabling, especially with great tools like VituixCAD where its immediately obvious what can be EQd and what not :)

ps. your listening window response looks ragged for tweeter part, I suggest to compromise the on axis response and find nice of axis response instead (5-20deg), eq for smoother listening window and power response for better sound. On-axis contains all acoustic issues of the construct and is not good candidate for optimization, unless acoustic issues are taken care of. Or perhaps you already have.
 
Last edited:
OP
voodooless

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
That doesn't look half bad. How much is that baffle curvature responsible for this?

It's interesting how accurate the datasheet is:
1657874233032.png

Here I overlayed the datasheet with the corrections. Pretty impressive. Have any polar plots?
 

Clavius

Member
Forum Donor
Joined
Jul 28, 2020
Messages
25
Likes
108
Location
Stockholm, Sweden
Look for VituixCAD measurement manual. If you perform messurements and convert impulse responses into frequency responses keeping the windowing same between all files, like the manual suggests, the timing info is preserved in the data and you dont need to fiddle with acoustic offsets and related issues, they are already included in the data.

Perhaps you knew this, still wanted to make the remark for future readers.

And yeah, DSP has far more positives for sound quality than negatives, even counting in cost of extra amplification and cabling, especially with great tools like VituixCAD where its immediately obvious what can be EQd and what not :)

ps. your listening window response looks ragged for tweeter part, I suggest to compromise the on axis response and find nice of axis response instead (5-20deg), eq for smoother listening window and power response for better sound. On-axis contains all acoustic issues of the construct and is not good candidate for optimization, unless acoustic issues are taken care of. Or perhaps you already have.
Yes, that’s exactly what we’re doing.
You can pretty much forget on-axis resp for most coaxes. There is also a problem with pinched directivity at 2.5k-3k that is being looked into. The real challenge is the passive filter though, unfortunately we need that because of commercial demand.
 
Last edited:

Clavius

Member
Forum Donor
Joined
Jul 28, 2020
Messages
25
Likes
108
Location
Stockholm, Sweden
That doesn't look half bad. How much is that baffle curvature responsible for this?

It's interesting how accurate the datasheet is:
View attachment 218383
Here I overlayed the datasheet with the corrections. Pretty impressive. Have any polar plots?
I do but can’t share too much at this point, sorry fellow shareholders and all that☺️. What I can tell you is that that it’s a passive cardioid and the bass is an aperiodic vent design. We will be joined up with the Klippel after summer vacations, hope to be able to share final active and passive measurements then!
 
OP
voodooless

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
I do but can’t share too much at this point, sorry fellow shareholders and all that☺️. What I can tell you is that that it’s a passive cardioid and the bass is an aperiodic vent design. We will be joined up with the Klippel after summer vacations, hope to be able to share final active and passive measurements then!
Nice! I've been thinking about something like that with a coaxial. Cool that it's being done! Mainly interested in the directivity of the HF part though. Is it even halfway smooth?
 

Clavius

Member
Forum Donor
Joined
Jul 28, 2020
Messages
25
Likes
108
Location
Stockholm, Sweden
Not exactly smooth, no..;) It should be noted that this is on an earlier baffle than the one I posted previously, and it has some diffraction problems due to how the driver was mounted. This one shows the response optimized to 10 degrees btwy. There are also some user errors (by me) done when exporting the Klippel file to .txt.
 

Attachments

  • A-Ha a2aa_d_KLIPPEL Directivity (hor).png
    A-Ha a2aa_d_KLIPPEL Directivity (hor).png
    66.5 KB · Views: 150
OP
voodooless

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
Not exactly smooth, no..;)
Not quite, but also not do bad. It's a bit of a mess around 2.5~3 kHz as we've already seen in the other graphs. That's about the wavelength of half the woofer size.
 

Plcamp

Addicted to Fun and Learning
Joined
Jul 6, 2020
Messages
860
Likes
1,315
Location
Ottawa
I guess I remain unconvinced that a large coaxial’s point source advantage is worth the deterioration of diffraction from edge effects, discontinuities in audio path vs an mtm arrangement of conventional drivers.

Most convincing for me seems to be coaxial compression drivers in horns covering from 500 hz up, where I can’t really see disadvantage, apart from cost. That delivers both point source and clean diffraction?
 
OP
voodooless

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
Most convincing for me seems to be coaxial compression drivers in horns covering from 500 hz up, where I can’t really see disadvantage, apart from cost. That delivers both point source and clean diffraction?
It's still not a point source with the woofer below 500 Hz not at the same axis, but close. Also, a 500 Hz horn is massive :eek:, more massive than a 15" coaxial. But yes, such a setup would also give a very nice result. I'm always in dubio between the two (I'd settle for 800 Hz crossover as a compromise, don't need a coax CD then). But for some reason, I get drawn back to these coaxes over and over. It's rather annoying ;).
 

Rednaxela

Major Contributor
Joined
Mar 30, 2022
Messages
2,052
Likes
2,675
Location
NL
[…] vs an mtm arrangement of conventional drivers.
Intrigued by this too.

Not wanting to go off topic too much, but what would be an example of a good implementation of this, comparable to what is being discussed here?
 

Clavius

Member
Forum Donor
Joined
Jul 28, 2020
Messages
25
Likes
108
Location
Stockholm, Sweden
A 15” crossed over to a BMS 4592ND on a really good horn at around 600ish would make for a very nice package but it will come with a form factor that will limit it’s appeal. A coax CD:s like the 4592 definitely needs time alignment to sound right though.
 

Jukka

Active Member
Joined
Apr 24, 2019
Messages
248
Likes
168
Here in Finland we got a designer who posted three size versions (8, 12, 15) of a 2-way PA coax: Piste. With this you need to pay attention to the CD, because it's quality can vary between coax elements. This is many years old, today I'd examine 18 sound drivers.

We also have Taipuu 3-way (scroll down), which has 15" coax with a 18" woofer for those who want go completely crazy. It uses BMS drivers.
 

Plcamp

Addicted to Fun and Learning
Joined
Jul 6, 2020
Messages
860
Likes
1,315
Location
Ottawa
Intrigued by this too.

Not wanting to go off topic too much, but what would be an example of a good implementation of this, comparable to what is being discussed here?
I am specifically interested in 5” or 6” sb acoustics ceramic drivers (that have superb distortion performance between 300 and 2.5 kHz) surrounding a waveguided dome tweeter crossing as low as I can…on a narrow curving backwards baffle.
 
OP
voodooless

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
I am specifically interested in 5” or 6” sb acoustics ceramic drivers (that have superb distortion performance between 300 and 2.5 kHz) surrounding a waveguided dome tweeter crossing as low as I can…on a narrow curving backwards baffle.
MTM with waveguide is actually not so easy. You’ll want to keep the center to center distance as low as possible, and a waveguide needs space.
 

mmi

Active Member
Joined
Jan 15, 2021
Messages
161
Likes
180
Intrigued by this too.

Not wanting to go off topic too much, but what would be an example of a good implementation of this, comparable to what is being discussed here?

Also hoping this isn't off topic and is what you are wanting to see. Well implemented d'appolito (there are also towers):

 

Plcamp

Addicted to Fun and Learning
Joined
Jul 6, 2020
Messages
860
Likes
1,315
Location
Ottawa
A low xmax 5” in an MTM w a dome is a different bowl of wax from a 10-12” pro coax driver with a cd though.
It is, it’s a radically different approach, with the inherent disadvantage that you can’t achieve true point source because you can’t cross low enough to be inside 1/4 wavelength driver separations. And, of course, twin 5” limits max spl to much less than a larger coax could deliver (sb drivers I am looking at limit on Klippel to 3 mm xmax).

Coax, big horn or mtm. I flip flop on this all the time and can’t make a decision. One day I will.
 

Jukka

Active Member
Joined
Apr 24, 2019
Messages
248
Likes
168
It is, it’s a radically different approach, with the inherent disadvantage that you can’t achieve true point source because you can’t cross low enough to be inside 1/4 wavelength driver separations. And, of course, twin 5” limits max spl to much less than a larger coax could deliver (sb drivers I am looking at limit on Klippel to 3 mm xmax).

Coax, big horn or mtm. I flip flop on this all the time and can’t make a decision. One day I will.
I've been at this crossroads before. MTM would require very small drivers to accomplish point source and then it would become a 3-way if you add a woofer. If going 3-ways including MTM, it would be a very big speaker.

Coax? It also needs some diameter and PA coaxes usually don't have very good CD with them. You could upgrade, but the wave guide (midbass part) is still moving, which will modulate high frequency response.

Coaxial compression driver? This beasts are capable of ~120 dB/W/m, but require a (very) large horn when going 500 Hz and below. And you would still need a woofer, so it's a three way.

My solution was to go with a CD in a horn with a 10" woofer, where crossover point can go ~1500 Hz, which is somewhat point sourceish, although not a desktop size speaker. -EDIT- I also got subs for the low end, so the complete system looks like a 3-way.

The new 1.4" CDs can go from 1 to 20 kHz, added with a suitable woofer these would be my choice design. Only downside is the price, those CDs are expensive.
 
Last edited:
OP
voodooless

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
The new 1.4" CDs can go from 1 to 20 kHz, added with a suitable woofer these would be my choice design. Only downside is the price, those CDs are expensive.
Of 1kHz crossover you don’t need a 1.4” driver especially not for HiFi. Quite a few decent 1” drivers will do that just fine at the added benefit of better HF. They are also less expensive.

I also doubt that the higher end PA coaxes have lower quality compression drivers. BMS offers coaxial CD’s, Lavoce shared tech with their high-end CD’s, Faital has annular CD’s as well as some larger ones of excellent quality.
 
Last edited:

gnarly

Addicted to Fun and Learning
Joined
Jun 15, 2021
Messages
990
Likes
1,390
Coax, big horn or mtm. I flip flop on this all the time and can’t make a decision. One day I will.
Yep, been the same decision set for me...although I have to add synergies/MEHs to the group.
Here's my 2c journey through those...

MTM, and my foray into DIY started with the PM90/60 as found on soundforums.net
They use a coax CD on a horn, and a pair of horn loaded 12"s wrapped around the CD/horn.
Truly great sounding/great measuring design. DSP required, meant for 100 Hz up.

pair.jpg

Then i went onto a modular MTM tangent, which put each driver in its own trap box, for stacking and arraying according to SPL/audience size needs.
Meant for PA, but came out surprisingly hi-fi like....but clearly not quite as clean as a single PM90/60.

After building those I kinda wandered around, still using the same compression drivers as in the PM90/60 (bms4594he and b&c dcx464) on various larger horns, trying go as low as possible.
Horns were b&c me464, jbl 2384, and a 31" DDS from a PAS box.
All of them worked well down to between 300-500Hz; I used 18" and 15" subs to reach up the xover points.
Very good sound, but ultimately not as coherent/convincing as the PM90/60s using same subs crossed at 100Hz.


The one large coax I tried, the b&c 15cxn88, I already posted about in this thread.
Twas fine, but it had the lowest level of performance of all I'd tried so far. And I read Tom D say their well regarded SM-80, which uses a 12" b&c coax, was not one of their more revealing speakers.
So I ditched further pursuit of a large coax. I think what KEF does with smaller ones may be where they work best. But that's pure speculation.

Then on to synergies.
All i can say is hell yes. It's funny, I can't get them to measure as cleanly as the PM90/60s, or even sometimes as clean as the big horn setups , but their clarity and overall coherent sound is my ticket/obsession.
All my Synergy builds have used the same CDs again, and have range from 31" to 48" wide, patterns from 60x40 to 90x60 with several in between. And all have been designed to cross to a sub at 100Hz.

Oh, with regards to horn pattern control...i just use Keele's classic formula: 1,000,000 = coverage angle in degrees x frequency x width or height in inches.
I've found it's not too far off for direct radiators, when the degrees used come from the ka charts.
So iow, I don't think a 500Hz horn has to be very much larger than say a 15" driver, to have the same pattern control at 500Hz.
 
Last edited:
Top Bottom