Hi all,
I'm starting a new project a little bit crazy (need to keep busy ^^).
There would be 2 speakers + 1 compression driver.
For the design, I was thinking about something like that:
Dimensions are 50x35x20cm approximately. With the compression driver mounted on a waveguide like this one:
The two mid/bass will be located within the horn walls.
Regarding the opening angle of the horn, I haven't decided anything yet but I was thinking something around 90 degrees. To be confirmed with a simulation.
For the low/mid section, I've selected 3 drivers that could suit: FaitalPRO 4FE32-4, 5FE100-4 and 6FE200-4.
I've done the first simulations with Hornresp (7L for the 4" and 5", 10L for the 6").
The front face of each speaker will radiate in a kind of horn, more a duct in fact (circle equal to Sd, 3cm deep). This explains the (wanted) ascending slope and the resonance at 1kHz.
I'll choose the driver size according to mechanical constraints (not sure if the 6 inch driver would fit ^^).
And I might completely change the dimensions because the volume is too big (15.4L but the drivers need only 8-10L).
In light of the above come some questions
:
And the filtering between boxes will be active.
I'm starting a new project a little bit crazy (need to keep busy ^^).
There would be 2 speakers + 1 compression driver.
For the design, I was thinking about something like that:
Dimensions are 50x35x20cm approximately. With the compression driver mounted on a waveguide like this one:
The two mid/bass will be located within the horn walls.
Regarding the opening angle of the horn, I haven't decided anything yet but I was thinking something around 90 degrees. To be confirmed with a simulation.
For the low/mid section, I've selected 3 drivers that could suit: FaitalPRO 4FE32-4, 5FE100-4 and 6FE200-4.
I've done the first simulations with Hornresp (7L for the 4" and 5", 10L for the 6").
The front face of each speaker will radiate in a kind of horn, more a duct in fact (circle equal to Sd, 3cm deep). This explains the (wanted) ascending slope and the resonance at 1kHz.
I'll choose the driver size according to mechanical constraints (not sure if the 6 inch driver would fit ^^).
And I might completely change the dimensions because the volume is too big (15.4L but the drivers need only 8-10L).
In light of the above come some questions
- I want the total impedance of each box to be close to 8 ohms in order to be able to use 2 boxes for one amplifier channel.
So I can use either 2x 4 ohms in series, or 2x 16 ohms in parallel.
Is there any advantage to one or the other solution?
In terms of acoustic design, I don't really want tu use 16 ohms speakers because they exhibit a bump around 100-300 Hz due to high Re.
- For the compression driver, I've selected several models : Oberton D2525, Beyma CD1Fe, B&C DE10 and BMS 4524.
Is there any ref that you would recommand to me and/or experiences with one of the above refs?
And the filtering between boxes will be active.
1) speakers with pattern flip always sound 'weird' to me, and it's not possible to fix them with EQ because the problem is the polar pattern
2) If you need a tweeter with narrow vertical directivity, ribbons and AMTs are cheaper than compression drivers and sound better. The advantage of a compression driver, for a line array, is maximum output. If you're listening at home, you probably don't need to generate 130dB.
2) If you need a tweeter with narrow vertical directivity, ribbons and AMTs are cheaper than compression drivers and sound better. The advantage of a compression driver, for a line array, is maximum output. If you're listening at home, you probably don't need to generate 130dB.
Hi Patrick,
Thank you for your answer.
I don't really get what you mean by pattern flip?
Do you mean the MTM arrangement?
It's not gonna be used for home listening but more for PA, voice, live music.
That's why I'm choosing compression driver 😉
Edit: found an answer about pattern flip there -> Understanding Horn Directivity Control
Thank you for your answer.
I don't really get what you mean by pattern flip?
Do you mean the MTM arrangement?
It's not gonna be used for home listening but more for PA, voice, live music.
That's why I'm choosing compression driver 😉
Edit: found an answer about pattern flip there -> Understanding Horn Directivity Control
Last edited:
Ribbons and ribbon-shaped compression drivers have a vertical directivity that's basically zero.
Because of their shape, when the wavelengths are shorter than the height, the directivity will be close to zero. For instance, a 4" tall ribbon will have zero degrees of beamwidth down to about 3khz.
But after that frequency, the beamwidth expands dramatically and nearly instantaneously. The net effect is that music and voice sound weird, because the pattern is wildly asymmetric.
I've built a ton of horns for car audio, and this asymmetry is the reason that most people get sick of horns in cars imho.
In a commercial line array, the elements are stacked vertically, which dramatically increases how low the array can maintain beamwidth control.
For instance, if you use a FORTY inch ribbon instead of a FOUR inch ribbon, the frequency goes from 3,375hz to 337hz.
In a line array, the narrow beamwidth is a "feature" not a "defect", because it prevents them from interfering with each other.
Basically there's no good reason to use a single element line array IMHO. If you don't need huge SPL, go with a conventional two-way. I have some Yamaha DXR 12s, those are nice for that.
Because of their shape, when the wavelengths are shorter than the height, the directivity will be close to zero. For instance, a 4" tall ribbon will have zero degrees of beamwidth down to about 3khz.
But after that frequency, the beamwidth expands dramatically and nearly instantaneously. The net effect is that music and voice sound weird, because the pattern is wildly asymmetric.
I've built a ton of horns for car audio, and this asymmetry is the reason that most people get sick of horns in cars imho.
In a commercial line array, the elements are stacked vertically, which dramatically increases how low the array can maintain beamwidth control.
For instance, if you use a FORTY inch ribbon instead of a FOUR inch ribbon, the frequency goes from 3,375hz to 337hz.
In a line array, the narrow beamwidth is a "feature" not a "defect", because it prevents them from interfering with each other.
Basically there's no good reason to use a single element line array IMHO. If you don't need huge SPL, go with a conventional two-way. I have some Yamaha DXR 12s, those are nice for that.
What is the desired coverage angle? Is subwoofer support available?
2x 16 ohms in parallel is preferable to 2x 4 in series, because amplifiers are voltage sources and drivers expect to see a certain voltage. If the drivers are not identical due to manufacturing tolerances, their impedances will differs and they will see different voltages when connected in series. When connected in parallel, they still see the same voltages.
Different impedance versions of the same driver should not differ in how much of a hump they give at 300 Hz.
2x 16 ohms in parallel is preferable to 2x 4 in series, because amplifiers are voltage sources and drivers expect to see a certain voltage. If the drivers are not identical due to manufacturing tolerances, their impedances will differs and they will see different voltages when connected in series. When connected in parallel, they still see the same voltages.
Different impedance versions of the same driver should not differ in how much of a hump they give at 300 Hz.
Last edited:
Patrick, as beamwidth narrows, directivity increases. A directivity of 1 is omni. I really have no idea what you were trying to say with a directivity of 0. On a dB scale, a direcitivity of 0 dB is omni.
Also, there's no abrupt change in directivity when the wavelength of sound equals the height of an array. Plus, uniform radiators like a ribbon always have sidelobes so they will always interfere with each other to some degree. (At certain frequencies, their sidelobes will be in non-acoustic space, which we don't care about).
I said "But after that frequency, the beamwidth expands dramatically and nearly instantaneously. The net effect is that music and voice sound weird, because the pattern is wildly asymmetric."
In other words, when the wavelengths become larger than the ribbon itself, the beamwidth increases.
Just to clarify a bit my "needs". Because it was not really clear the first time ^^ Sorry :/
I intend to use those boxes by group of 4 at list for each side.
The desired horizontal coverage angle for each array is 90-100 degrees.
Yes I already have some dual 18 subs to cover the LF region.
Regarding the difference between 2x4ohms in series (gray curve) and 2x16ohms in parallel (black curve), this what I was talking about.
Here the bump is around 200Hz for the 6FE200.
I intend to use those boxes by group of 4 at list for each side.
The desired horizontal coverage angle for each array is 90-100 degrees.
Yes I already have some dual 18 subs to cover the LF region.
Regarding the difference between 2x4ohms in series (gray curve) and 2x16ohms in parallel (black curve), this what I was talking about.
Here the bump is around 200Hz for the 6FE200.
- Status
- Not open for further replies.