Dispersal pattern with WWTWW?

[Fusion916]

Considering building a 2 way "line array" in the following configuration:

W
W
T
W
W

With the following drivers:

http://www.parts-express.com/pedocs/...ifications.pdf
http://www.parts-express.com/pe/show...number=276-400

Crossed at 4k @ 24db

The sensitivities math (96db and 97db), but because there are 4 woofers and 1 tweeter, is there going to be an "apparent" different in volume and dispersion past say 8k or so?

[Sydney]

The sensitivities math (96db and 97db), but because there are 4 woofers and 1 tweeter, is there going to be an "apparent" different in volume and dispersion past say 8k or so?

When drivers are physically separated ( the W's in your example ) the distance between them causes dispersion pattern interference ( constructive and destructive ) . This restricts the bandpass of mutual coupling and increases directivity. So the same xover used for a single MT is not appropriate for a WWTWW

[Fusion916]

When drivers are physically separated ( the W's in your example ) the distance between them causes dispersion pattern interference ( constructive and destructive ) . This restricts the bandpass of mutual coupling and increases directivity. So the same xover used for a single MT is not appropriate for a WWTWW

So the xover points of the tweeter and woofer have to be different? Do you have an article or anything explaining this in greater detail?

[curt_c]

There will be a few different things happening as the frequency increases:

By 4K the Aura drivers will be exhibiting significant beaming, and dispersion will be less then optimal with an off axis flare developing above the crossover frequency. They may sound a bit reticent in the midrange and lower treble.

As Sidney states: Above about 1500 Hz the vertical response of the inner woofers will exhibit comb filtering off axis. The outer woofers even earlier, but the filtering may be ameliorated by the number of woofers. They will be best listened to at the tweeter height.

There is a pretty significant peak of about 5 dB, -likely a cone break up mode, in the Aura driver above 3K that will not improve the sound if crossed at 4K. Likely it will sound as if the tweeter is harsh, when it is really the woofer.

Don’t forget baffle step. Unless these are going to be in walls, the design will likely need at least 3 dB of baffle step compensation, making the apparent sensitivity of the series’d, paralleled, woofers at around 90-91 Db/2.8v/1m.–The speaker gods giveith, and the speaker gods taketh away… In this case, the ribbon will require attenuation.

C

[greywarden]

You should do a 2.5-way

[scottq]

The lower frequencies won't lobe as much due to driver spacing, but the mids would start becoming a problem... not to mention the tweeter interaction. So, I would suggest running the outside woofers as low frequency only, and use the inner woofers as mids - something like a Woofer Midwoofer Tweeter Midwoofer Woofer configuration. You'd then get something similar to a MTM/D'Apolito dispersion pattern (assuming all vertically aligned).

[Sydney]

So the xover points of the tweeter and woofer have to be different? Do you have an article or anything explaining this in greater detail?

Lower...
Note what Curt said:

Above about 1500 Hz the vertical response of the inner woofers will exhibit comb filtering off axis. The outer woofers even earlier, but the filtering may be ameliorated by the number of woofers. They will be best listened to at the tweeter height.

Because sound wavelengths vary, the effect of distance of separation varies with frequency.
I don't have a simple single article: It involves an individual speakers Ka, and takes into consideration what the difference in distance from sources means in terms of arrival time delay
Use Google book preview on this title:
Loudspeaker and Headphone Handbook edited by John Borwick
See Section 14.1 page 16
Pattern Interference is the operating principal behind line array design. CBT designs compensate for the issues mentioned prior.

[Fusion916]

What are you basing the cone breakup frequency off of? These won't be infinite baffle but I don't mind making the boxes relative big so wouldn't there no attenuation of the woofers?

[Fusion916]

The lower frequencies won't lobe as much due to driver spacing, but the mids would start becoming a problem... not to mention the tweeter interaction. So, I would suggest running the outside woofers as low frequency only, and use the inner woofers as mids - something like a Woofer Midwoofer Tweeter Midwoofer Woofer configuration. You'd then get something similar to a MTM/D'Apolito dispersion pattern (assuming all vertically aligned).

Doesn't this lose the SPL gain from multiple woofers though?

And are any of these issues eliminated by increasing the spacing between woofers?

[Sydney]

No free lunch and increasing spacing makes matters worse

[Fusion916]

Would be better to cross lower? Such as 3K? And would an EQ solution be better? I don't really want to go 2.5 or 3 way. Even with even sensitivities, I don't want to run into the scenario where some lower tones (~2k and below) sound "big" and higher tones (maybe 6k and above) sound "small" due to the majority of the sound coming out of 2 speakers instead of 8.

[Chris Roemer]

You need to start reading some basic speaker design theory.

It's usually more cost effective to learn first, build second, design third.

[fbov]

Go here:
http://speakerdesignworks.com/index_page_2.html

Read about the first two designs listed, Stentorian and Sten II. The re-read post #4 with the understanding that Curt's used this driver and so knows where of he speaks.

In fact, the biggest thing for you to reallze is that Curt's already done the hard part; you should consider building one of Curt's...

HAve fun,
Frank

[r-carpenter]

What are you basing the cone breakup frequency off of? These won't be infinite baffle but I don't mind making the boxes relative big so wouldn't there no attenuation of the woofers?

cone break up mode relates to the half size of the cone diaphragm. That is the main mode. Often, dust cup will have a break up as well or there are some issues between the back of the diaphragm and frame/spider. They are all inherent qualities of a driver. Regardless of a baffle placement, they will stay there. Only thing you can do is address these issues in the crossover and limit pass band of a driver. Basically, you need to roll off the response before it get's to the problematic areas.

[Fusion916]

I did a little bit of research on cone breakup last night and from what I understand it's purely frequency dependent for the driver (baffle has nothing to do with it). From the frequency response graph for the Aura drivers, it seems like cone breakup occurs slightly after 2k, is this correct? Meaning i need to xo the drivers at 2k or below if I want to alleviate the cone breakup issue?

[r-carpenter]

there are some issues that this Aura has after 2k but the main cone resonance is around 5k. You can see it as a sharp bump in a frequency response graph. 5k-2.7" half size of a diaphragm.
You are not likely to eliminate 2k issues also because AC tweeter needs to be crossed higher then that. And it is somewhat directional vertically. I am almost ready to suggest a different tweeter or a driver combination/configuration. Previously suggested 2.5 sounds like a good idea as well.

[Fusion916]

Well I already bought 16 of them (buyout price) so I will be using them for something.... I might just XO at 2.5k at 24db and pray.

[r-carpenter]

I'd use them (aura) with robust dome tweeter and try to bring crossover point as low as possible. Dayton RS28 aluminum dome you could probably bring in to 1800hz or so and avoid a multitude of issues. Not impossible to use ribbon here, just not ideal and rather difficult even for a seasoned designer.

[scottq]

Doesn't this lose the SPL gain from multiple woofers though?

And are any of these issues eliminated by increasing the spacing between woofers?

To be clear, I was essentially discussing an MTM 2.5 way...

And, it's a complicated answer w/r/t sensitivity. It all depends on your initial wiring plans. Wiring 4 8-ohm drivers in parallel is going to give you a nominal 2ohms. Most amps aren't going to like that one bit. But, yes, it would give you a huge boost in sensitivity for the frequencies those drivers cover.

If your amp will handle 2 ohms, you could run all 4 in parallel. Put a low pass on the two outside woofers at a low frequency, as would be done with the 0.5 on a 2.5 way. Then run a normal low pass on the two midwoofers as you would on the low frequency driver of a 2.5 way. This would still give your "midrange" a ~6dB boost, and your "bass" would get about a 12dB boost... but, if you factor in baffle step loss, the "bass" would then approximately match the 6dB boost of the mids. With this, it would only be "2 ohms" at low frequencies... as it transitions to "midrange" and the other woofers are cut out, impedance would increase to 4ohms as the inner woofers would be the only ones contributing. If you have a monster amp that's stable at 2ohms, this could work.

If you're going to do series/parallel combo wiring to maintain 8ohms on the "woofers" (which would then be and 16ohms on the "midrange"), things get more complicated. Envision having two 8ohm woofers in series, and two 8ohm mids in series... then, put the woofers (16ohms) and mids (16ohms) in parallel with each other. At low frequencies, you'd get a 6dB boost (compared to 1 driver), which could compensate for baffle step losses, and you'd be at 8ohms at low frequencies. At mid frequencies, you'd be back to 0dB gain (compared to 1 driver), and at 16ohms in the mids.

I think that's right - hope it makes sense...

[Fusion916]

Just to be clear, I am going to do a series/parallel combination, 4 woofers per speaker, for a nominal 8 ohm impedance. Not going to mess with the issues of going 2.5 way as I am a newbie designer.