What about an adapter to use a 1.4" driver on the PH-2380? E.g. the HF146R has a exit angle of 29 degrees; the only information about the throat angle of a JBL 2380a mentions roughly 25 degrees.
Is it as 'simple' as matching the diameter and angles on both side with a smoot as possible transition?
Is it as 'simple' as matching the diameter and angles on both side with a smoot as possible transition?
The HF206 is a HF146 Driver with an adapter to make it have a 2" exit, that can be removed so it is quite flexible.
Watching video's by Toole and Geddes have pointed my towards (constant) directivity. You guys mentioned it also, but it took some time for me to sink in. My current understanding is that both the on and off axis response should be quite flat for it to sound good, which is reflected in constant directivity. On axis response can be made flat with a dsp. Weird behaviour in the off axis area cannot be fixed (at the same time).
Looking for constant directivity horns I found the XT1464C. Its datasheet (https://www.eighteensound.it/media/W1siZiIsIjIwMTkvMDcvMTkvMTBfNTdfMzlfMzgxX1hSMTQ2NEMuUERGIl1d) shows a fairly constant directivity of 10dB above ~1.1kHz.
I found a plot for the directivity of a woofer (https://greatplainsaudio.com/wp-content/uploads/2017/07/TL_237.pdf). I also found an excel file that calculates it (http://www.tonestack.net/software/speaker-directivity-simulators.html v1.03)
12" woofer:
A 15" woofer has a directivity of 12-13dB at 1.1kHz, a 12" is at 10-11dB. I don't have a feeling for these numbers, but a 12" seems a better match.
Have I discovered what you already told me?
You seem have, at least the root of it!
After you read more constant directivity is actually quite hard to achieve, for wide bandwidth at least. Full bandwidth almost impossible because lows are naturally very omni and highs like to beam interact with stuff, narrow beam. Usually the best thing to do, with almost any system, is to have smooth directivity without sudden changes. This means the reflected sound in the room is roughly the same (frequency response) as direct sound which seems to make loudspeakers rank better in blind listening tests. Note, omni speaker would also have constant directivity. Narrower directivity which you can achieve with waveguide for example, reduces early reflections in the room, which should increase "clarity", depending how you position the speakers in the room of course.Most of the sound you hear in a typical far field listening situation at home is reflected sound, only small percentage is direct sound. Some threads I read late yesterday wrote there is about 20/80% ratio for direct / reflected sound at listening spot, which can be calculated from listening window and power response or something like that, with some assumptions, don't have links at hand. Look it up if interested. With constant (or smooth) directivity you can EQ the listening axis response to flat and the off-axis sound is also EQ:d to flat or at least smooth slope. Sloping down "off-axis", alias sound power is perceived be best. If it was flat many say it is too much highs. Hence, constant directivity might not be the best for all situations (listening environment). Perhaps the middle frequencies benefit most of constant directivity and perhaps the highs can beam a bit without issues and is "easy" to actually make. One of the main things is to choose good waveguide for tweeter and then matching sized woofer for it in order to have their directivities narrow roughly on same bandwidth, where you would put the crossover. This yields smooth DI around the crossover.
While you get deeper into it, check what you can do with early reflections, and sweet spot!
Last edited:
See post 5 lines 7,8,9. A 15" woofer + 7"x17" horn crossed at 1200 hz maintained +-3db evenness over 22.5 deg from straight ahead. 500 hz to 16 khz. Peavey SP2-XT. See also Pispeaker 4 pi referenced above. Directivity industry wide is determined -6 db, I hear differences of 3 db so I'm more interested beam witdth at that level of flatness. Using 12" instead of 15" woofer means 54-100 hz is difficult to maintain - 3db. Crossing above 1200 with 15" means some beaming as you see in your post. Some CD drivers will down go to 1200 hz or even 800 hz with significant watts rating. Going 3 way also stop the mid-range beaming, but with 70 & 100 watt compression drivers available that go to 500 hz, not necessary. 3 way introduces more sound modifications in crossover and much more cost (passive) in coils & MKT caps than a 70 W CD and a one point passive crossover. 2 way 12 db passive crossover is 2 coils, 2 capacitors, 2 or 3 tweeter level (reduction) and protection parts. With 54 hz - 3db, I live without a sub amp consuming electricity 14 hours a day.
I did some tests to prove/disproove theory that backing speaker up to hard wall stops bass droop of the box itself. My SP2(2004) drops about 20 db from 54 hz to 30. In my music room backing op to hard wall reduced droop to about 6 db. Response below 30 hz is a low hum from some diesel sewer pump somewhere. Left chart is back of speaker against wall, 2nd is one foot from wall. 3rd is 3 and half foot from wall.
Attachments
Last edited:
Keep in mind that big woofers break up and that partial cone movements will influence directivity. Normally, you would measure directivity by acquiring on- and off axis measurements. Next step is picking the waveguide or horn plus driver and acquire the same measurement set. Last step is matching both sets, which gives you a clue about the desired crossover frequency. Plus you have to determine what vertical behavior you want to establish, which narrows the search for the crossover frequency and topology. Throw into that the requirements of the drivers (keeping away from distortion) and presto! you are close to a design goal.
EQ is 1 dimensional so it will apply to all the curves if applied to the whole speaker that is why simulating in Vituix with full polar data shows the effect on all of the curves simultaneously.
Earl advocates for a flat constant DI as high and low as it can reasonably be achieved, which in his speakers is the waveguide and a small region at crossover to the woofer.
Toole suggests flat on axis smooth off axis tending towards constant directivity.
Constant directivity would have the off axis curves be the same as the on axis just at a lower SPL. To truly get this over a wide bandwidth can be done but often comes with other tradeoffs, dipole and cardioid. When the DI is flat in an axisymmetric waveguide there tends to be some disturbance of the 0 deg on axis and near axis curves. There is usually a smooth response in the 10 to 25 degree curves which if taken to be the on axis by toeing the speaker in give a nice listening window of very similar on and off axis results.
Allowing the DI to rise gradually as frequency rises gives the opportunity to have a smoother response to all curves without the disturbance noted above. There is no proof that this is better or worse or that any particular directivity has been shown to be preferred. What seems to matter most is getting in the ballpark.
Having the directivity indices match is one part but that does not consider the pattern width at the crossover frequency. A polar map of the waveguide will show the width of the pattern. Vituix can simulate the piston size of the driver on the intended baffle and then you can see where the directivities and patterns match the best. Two waveguides could have the same DI but different patterns.
Seems like your more concerned with picking between a 15" or 12" woofer at this point. If either woofer is expected to play up into the mid range frequencies, you're going to want one that does the best job up there ... not so much how low you can push it. In fact, a woofer that plays well in the mid won't go very low (curvilinear coned). You can push it but the mid range starts to suffer in quality, to me at least. A 15" woofer can play well in the mid range ... say even to 1khz if you had to and still get down to 45hz with no problem (sealed). I doubt that a 12" can do that (sealed) at least not with the same authority. So a 12" would probably need a sub to satisfy. A 15" may or may not and that depends on your needs.
If this isn't for home theater duty I'd go with the 15" which can be crossed to a 1" compression driver around 900hz as Geddes has shown. The Faital 15pr400 (curvilinear cone) plays nice up this high (+ if needed) and gets low enough in a sealed box (180l) with a little EQ.
If this isn't for home theater duty I'd go with the 15" which can be crossed to a 1" compression driver around 900hz as Geddes has shown. The Faital 15pr400 (curvilinear cone) plays nice up this high (+ if needed) and gets low enough in a sealed box (180l) with a little EQ.
It's the stuff I have to buy. Currently I see two options: 15" + 2"CD + 2380 horn, or 12" + 1.4"CD + xr1464c horn.
I have to check availability and price. New drivers are expensive, second hand is much better, but availability is worse. I'm thinking I need to spend some more on horns as they dictate the directivity.
EQ and enclosure size can be determined later.
I have to check availability and price. New drivers are expensive, second hand is much better, but availability is worse. I'm thinking I need to spend some more on horns as they dictate the directivity.
EQ and enclosure size can be determined later.
2"CD may be unnecessary expense on a 15" woofer. Peavey is using 1.4" on their best sounding 15" product the SP2(2004). Pispeakers is crossing a 1" CD @ 1200 hz in the 4pi. The crossover point determines how much power goes to the CD and that means how loud you can play within the watts limit. A 70 W CD crossed at 800 hz should be able to drive you out of your 4 m room. AES watts rating are pink noise, apparently rather conservative. 15" woofers don't tend to break up until 2000 hz or higher, but they usually have a sound peak at 1200 to 1500 which makes that portion beam if you use it. I don't intend to. I don't need a 500 W (AES) speaker.
In a related issue, CD protection in the crossover is a money saver. Usually a incandescent 12 v dome light bulb series the tweeter. Pops from cable fallout or mike drops happen, cause huge treble spikes. I've blown a tweeter with a 35 w/ch amp.
I can't find a horn that has constant directivity to a frequency low enough where it matches a 15".
^ You could order custom made from Joseph Crowe for example, or from some other "small manufacturer". Or design your own and get manufactured (ATH).
The thing is, you probably want good treble as well since the system is for HIFI application. 2" is a strectch while 1" would be the best. 1" drivers would work just fine for home application when there is no need for big SPL like in public address situations, where the drivers are mostly specified for. So you are correct, problem is to find suitable waveguide.
The thing is, you probably want good treble as well since the system is for HIFI application. 2" is a strectch while 1" would be the best. 1" drivers would work just fine for home application when there is no need for big SPL like in public address situations, where the drivers are mostly specified for. So you are correct, problem is to find suitable waveguide.
The CD's I've seen have the resonance much higher, which was said to be a problem here. Or is that resonance peak dependent on the horn?