I think even Beyma's own datasheet shows quite high distortion at some places. Worst thing being it's not 2nd but 3rd harmonic that sometimes gets quite high.
I think next friday we might have some answers on what direction this effort will take. Or if it will... And there are indications this might become more than a one man effort. Seems I'm not the only one interested.
Currently the vague concept would be something along the lines of a 150-180 litres bass reflex with the B&C 15FCX76. Not sure yet if I'll pull the trigger on this.
The thing I find tough about Coaxial drivers is - where to set the crossover point?
The woofer cone rarely seems to make a horn quite big enough for the tweeter to cross at a point low enough to be inside the good response of the woofer cone. The smaller the cone, the higher you can cross it, but then the smaller the resulting horn! The two chase each other.
I suppose with DSP and very steep slopes it could be done neatly. I've had some luck with passive crossovers on 18" and 15" coax. Not so good on smaller.
The woofer cone rarely seems to make a horn quite big enough for the tweeter to cross at a point low enough to be inside the good response of the woofer cone. The smaller the cone, the higher you can cross it, but then the smaller the resulting horn! The two chase each other.
I suppose with DSP and very steep slopes it could be done neatly. I've had some luck with passive crossovers on 18" and 15" coax. Not so good on smaller.
I think this is going to be pretty easy in that respect. When you look at the compression driver response graph, you see that it "tops out" at around 1,2k, which looking at the directivity maps is quite a nice place to cross over from the woofers. An electrically first order filter around 8-10kHz should do the trick. Now, the woofer might need a little more work to get the slopes to agree, but I'm not really intimidated by that.
No, I mean the cutoff frequency. The response of the compressor falls ~6dB/octave with rising frequency, so a first order high pass centered this time at around 8-10kHz will tilt that response back to something resembling straight. Not always sufficient, but sometimes all that's needed. So the electric "crossover point" is totally different from the acoustic one. So when calculating the cap, it would be something like 1/(2*pi*8kHz*8ohm), roughly, and that combined with the real response of the driver gives a natural crossover point somewhere around 1,2kHz looking at the datasheet.
Why I use 8-10kHz? A rough guesstimation looking at the datasheet, that this is the point where the response stops rolling off. This will of course have to be implemented, measured and iterated.
Why I use 8-10kHz? A rough guesstimation looking at the datasheet, that this is the point where the response stops rolling off. This will of course have to be implemented, measured and iterated.
Neat.
An additional steeper HP filter slightly below xo point might give a better protection to the compression driver, though. This could be optional considering the application.
BTW, anyone remember this old thread?
http://www.diyaudio.com/forums/multi-way/132068-loading-tweeters-large-drivers-coaxes.html
An additional steeper HP filter slightly below xo point might give a better protection to the compression driver, though. This could be optional considering the application.
BTW, anyone remember this old thread?
http://www.diyaudio.com/forums/multi-way/132068-loading-tweeters-large-drivers-coaxes.html
Last edited:
OK, so I'm working to make this thread an exception to the previous rule, that all coaxial threads die before achieving a result. Still some things to sort out, but I'm decided on a path and shall post results, whether for good or for bad. This B&C driver just looks too promising to let go without at least giving it a go.
To give this some more weight, this is developing into a team effort instead of a one man exploration of a concept. We'll see what happens. Hopefully a wonderful loudspeaker.
To give this some more weight, this is developing into a team effort instead of a one man exploration of a concept. We'll see what happens. Hopefully a wonderful loudspeaker.
I like coaxial a lot - including Frazier's CAT40 - the speaker on the right would be my normal reference for a coax cabinet - its currently populated with a 15" neodymium P-audio coax - - oops - the "large format" part did not sink in - my coaxial are all small/1" format
An externally hosted image should be here but it was not working when we last tested it.
Last edited:
If you have good high frequency hearing, you will probably not be satisfied with a 2" exit compression driver, and just marginally staisfied with a 1.4" exit CD. There are 10" and 12" coaxials which use a 1" compression driver with a 36mm or 44mm diameter dome which offer 18Khz high frequency output.
For 15", look at the HF SPL plot for 3" diamater dome with 1.4" exit compression drivers.
You could build a separate 15" or 18" woofer box and use it with different T-M top boxes. A sealed box T-M will sound better than a ported alignmant. A 20" - 24" wide cabinet will push the baffle step compensation onto the woofer. VERY clever room/wall placement can smooth in-room bass without active equalization. Old school,
For 15", look at the HF SPL plot for 3" diamater dome with 1.4" exit compression drivers.
You could build a separate 15" or 18" woofer box and use it with different T-M top boxes. A sealed box T-M will sound better than a ported alignmant. A 20" - 24" wide cabinet will push the baffle step compensation onto the woofer. VERY clever room/wall placement can smooth in-room bass without active equalization. Old school,
It's true, there's a risk that the last octave will not be immaculately reproduced. If the B&C can manage it's stated upper limit of 18kHz, it should still be just enough.
The smaller compression drivers usually have problems being crossed low enough, and in my opinion that's a bigger problem than losing some of the last octave, where there is little information anyway. Upper midrange/lower treble dip caused by HF driver response falling of too high is much more disturbing.
When I auditioned those 18Sound 15-inch coaxials I could just hear, that they lose output beyond 14kHz, but it didn't bother me that much. The lack of energy between ~700-1500Hz did.
The smaller compression drivers usually have problems being crossed low enough, and in my opinion that's a bigger problem than losing some of the last octave, where there is little information anyway. Upper midrange/lower treble dip caused by HF driver response falling of too high is much more disturbing.
When I auditioned those 18Sound 15-inch coaxials I could just hear, that they lose output beyond 14kHz, but it didn't bother me that much. The lack of energy between ~700-1500Hz did.
Here's a response graph of that said Eighteen sound unit, measured by AudioVideo.fi:
The response under 300Hz isn't representative of real performance but an artefact of measurement setup. But here you can see the upper midrange/lower treble dip I was talking about.
An externally hosted image should be here but it was not working when we last tested it.
The response under 300Hz isn't representative of real performance but an artefact of measurement setup. But here you can see the upper midrange/lower treble dip I was talking about.
That dip is due to combined raw responses of the HF and LF drivers. The LF starts falling off at around 800Hz, and the HF driver can't go low enough to counter this. Would take a few LCR:s to fix that. Doable, of course. Still, rather choose a driver that doesn't need a truckload of stuff in the crossover to make it decent.
A 5dB dip spot in the middle of the most sensitive band of hearing isn't good. True, on some music you can't even spot it. But especially some vocals suffer badly. Though I can tell you that making vocals - especially the more intimately recorded - sound just right is a balancing act of itself, I struggled with this problem already with my smalle BMS coaxials, but in the end got it right.
A 5dB dip spot in the middle of the most sensitive band of hearing isn't good. True, on some music you can't even spot it. But especially some vocals suffer badly. Though I can tell you that making vocals - especially the more intimately recorded - sound just right is a balancing act of itself, I struggled with this problem already with my smalle BMS coaxials, but in the end got it right.
I don't know about those smaller ones. The one on the 12C362 sounds excellent! I really love it. Though currently I have it padded down a dB too much, but when it's right, it's dynamic, clean and realistic in a way I've never heard from a single dome. Then again, I haven't been much in to domes anyway for a while now.
But good news. I managed to sell some drivers I had laying around I didn't need anymore, so now I can confirm I'm going to pull the trigger on these B&C coaxials. We'll see how they'll perform in real life, maybe in just a couple of weeks.
But good news. I managed to sell some drivers I had laying around I didn't need anymore, so now I can confirm I'm going to pull the trigger on these B&C coaxials. We'll see how they'll perform in real life, maybe in just a couple of weeks.
I'm sure that's what they'd state. But the dip's wideness is what really makes it distracting.
So far I've heard the one on the 12C362 and a 4544 on a Beta 10CX. The latter also sounded very nice, but the shape of the Eminence's cone wasn't enough to bring the crossover point low enough.
And, by the way, I'm not building anything out of the 18Sounds, that was another project, featured in AudioVideo.fi. I had a chance to hear these, so I crossed that one out of my list of candidates. I just sent e-mail to a german supplier to order a pair of B&C 15FCX76's. Almost 800€ shipping included, I sure hope these will turn out right. They look promising, but you never really can't tell before trying them out. And even if their raw responses are good, how will they respond to a crossover?
Time will tell, but I'll keep updating this thread so that it won't turn out a dead end like all the others. I'm sure I'm not the only one on the planet on a look out for a decent 15" coaxial capable of reproducing a wide range enough not to require a sub. And which isn't a Tannoy...
And, by the way, I'm not building anything out of the 18Sounds, that was another project, featured in AudioVideo.fi. I had a chance to hear these, so I crossed that one out of my list of candidates. I just sent e-mail to a german supplier to order a pair of B&C 15FCX76's. Almost 800€ shipping included, I sure hope these will turn out right. They look promising, but you never really can't tell before trying them out. And even if their raw responses are good, how will they respond to a crossover?
Time will tell, but I'll keep updating this thread so that it won't turn out a dead end like all the others. I'm sure I'm not the only one on the planet on a look out for a decent 15" coaxial capable of reproducing a wide range enough not to require a sub. And which isn't a Tannoy...