Line Array Tweeter line length?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.

opc

Member
Joined 2004
Paid Member
Hi Jim,

Could the same end result (controlled horizontal dispersion) be achieved with a horn loaded design for the array?

I see a few potential issues with the symmetrical array:

1. You triple (at least) the cost of the array

If a 30 driver array is being planned, then you'll be looking at adding an additional 50 drivers or so to make a horizontally symmetrical array.

2. You triple the work involved

Even if you have the baffle CNC cut, it will still take months to wire everything up, mount all the drivers, and seal everything up such that there's access to the driver should one fail. At 4 screws per driver, you're looking at installing 640 screws for a stereo pair!

3. Crossover design will not be trivial

I found it extremely difficult to measure a line array in a domestic setting, which will make the development of a proper crossover very challenging.

4. Physical size is drastically increased

The baffle will now need to be significantly wider, pushing the drivers even closer to the side wall. A minimum 10-12" baffle would be needed to accommodate all the drivers.

With a properly designed horizontal horn, you'd get similar horizontal dispersion control, plus a significant gain in efficiency for minimal extra cost and work. The physical size could still be an issue, but if this is sitting on top of large subwoofers, then the physical size of the horns might fit just fine.

Cheers,
Owen
 
Symmetrical Line Arrays

Owen,

Answers to Questions:

1. "You triple (at least) the cost of the array. If a 30 driver array is being planned, then you'll be looking at adding an additional 50 drivers or so to make a horizontally symmetrical array.

I would say that typically you would use midrange drivers on each side of the tweeter line so it wouldn't triple the cost but it would be a cost increase vs. a typical two line array..

2. "You triple the work involved. Even if you have the baffle CNC cut, it will still take months to wire everything up, mount all the drivers, and seal everything up such that there's access to the driver should one fail. At 4 screws per driver, you're looking at installing 640 screws for a stereo pair!"

Line arrays are a lot of work so you would have more of everything to do.

3. "Crossover design will not be trivial. I found it extremely difficult to measure a line array in a domestic setting, which will make the development of a proper crossover very challenging."

I have found that outdoor measurements work except when the barking dog next door wishes to contribute to the design. Otherwise measurements in a large space--my two car garage qualifies.

4. "Physical size is drastically increased. The baffle will now need to be significantly wider, pushing the drivers even closer to the side wall. A minimum 10-12" baffle would be needed to accommodate all the drivers.
With a properly designed horizontal horn, you'd get similar horizontal dispersion control, plus a significant gain in efficiency for minimal extra cost and work. The physical size could still be an issue, but if this is sitting on top of large subwoofers, then the physical size of the horns might fit just fine."

Size does matter but you weigh size versus benefits. A horn (or these days a waveguide) would help the in-room directivity but at the expense of more a complex configuration. Essentially you have control of the horizontal dispersion with a horn/waveguide.

Another thought is to use a Tom Danley invented Paraline which might reduce the depth of the waveguide as claimed by Yorkville with their work with Tom:

Yorkville Sound: Paraline

Finally a more reasonable symmetrical array for a DIY'er is the Selah Audio array as shown.
An externally hosted image should be here but it was not working when we last tested it.


Jim
 
Thanx Jim. The Selah array shown is pretty much exactly what i had in mind, just a tad bit shorter at 48" or so and using the Vifa 2" drivers in leu of the ribbons. I haven't been involved in the commercial PA industry for 10 years now but controlled horizontal dispersion was desired in arrays back then and the woofer flanking design stuck with me as the best( but not the cheapest) way of getting that done. One only needs to listen to a horizontal MTM center channel to experience the effect.
 
An externally hosted image should be here but it was not working when we last tested it.


Jim

I've done this with 4" drivers - did not like it at all. :eek:

Finally (after a lot of crossover variations - freq. and order under a variety of baffle conditions), a .5 below 300 Hz was barely acceptable.

For lack of a better expression, I hear both lines when it's in the upper mid.s and find the result "confusing". It also seems to "push" imaging forward and reduce the perception of depth. Integration in even a large room really didn't "gel" until I was nearly out the door of the room, and long before then (despite its more directive nature) I was hearing far more "room" than any resulting offset from the added directivity.

Again though, that's just me. :eek:
 
Last edited:
Thanx Scott. While your ability to discrern the two lines is surprising, the rest of your obesrvations aren't as many don't care for constant directivity systems which may be the result of the two lines. I admit that such a design might be a bit much for my space.

I've ordered 24 of the two inch and 16 of the three inch for now( PE backorder so i'll have to wait a little while). I'll build two open baffle lines for now and take some measurements and try sub integration first. This should also allow me to integrate the two in a two way to say if there's any advantage to a first order high pass on the smaller drivers.
 

opc

Member
Joined 2004
Paid Member
Thanx Scott. While your ability to discrern the two lines is surprising, the rest of your obesrvations aren't as many don't care for constant directivity systems which may be the result of the two lines. I admit that such a design might be a bit much for my space.

I've ordered 24 of the two inch and 16 of the three inch for now( PE backorder so i'll have to wait a little while). I'll build two open baffle lines for now and take some measurements and try sub integration first. This should also allow me to integrate the two in a two way to say if there's any advantage to a first order high pass on the smaller drivers.

Very good idea on the prototype work! Although I agree with Scott on this one, we're well into the realm of personal opinion, so it's probably best to build a few variants and see what works best in your space with your own ears.

In general, I feel that the downsides and potential for error with introducing another line and a crossover, purely for the sake of horizontal dispersion control is really not worthwhile in a domestic setting. If this were a PA application and the coverage area requirements were very specific, then it seems to make more sense. In your home though, I think the added complexity and crossover point has too much of a negative impact on sound quality.

Try it either way though, and let us know what you decide!

Cheers,
Owen
 
Just a couple of comments:

Line arrays have no impact on horizontal directivity, which remains that of a single element. Putting side by side lines just adds a horizontal crossover/directivity aspect to the design. The main reason to do it, I would think, would be if you needed the greater bandwidth and power handling that you might pick up. It is definitely an added complication and will somewhat reduce the elegance of a full range array.

One thing to remember is that you pick up huge output at LF with a multielement array. For every doubling of units you get coherent summing and +6dB in level (less so at HF in the nearfield). So if you can find a wide range unit you can use it much lower in frequency than a single element would survive. Certainly full range arrays crossed over to subwoofers are a practical approach.

A lot of the fat roll inverted dome wide range units look very attractive in this application.

Regards,
David S.
 
Like the Aurasound NS3?.......lots of excursion but a rough top end.....would certainly need a tweeter line with those......which I'm not opposed to.

But I think you've uncovered my doubt....that neither the 2" or 3" Vifa's that I've mentioned will get to a sub crossover without significant HD. We'll certainly find out once the drivers I've ordered arrived. Thanx for the feedback Dave. I look fwd to getting further input from you once the proto's are up and running.
 
Line arrays have no impact on horizontal directivity, which remains that of a single element. Putting side by side lines just adds a horizontal crossover/directivity aspect to the design.

Regards,
David S.

Yes, any real "benefit" in increased horizontal directivity is over a fairly small bandwidth (..if it's there at all).

Worse, it's typically in an area where:

1. It's disadvantageous (from a psychoacoustic perspective), and
2. Where the tweeter line is often radiating as full omni - making integration that much harder at the crossover (..unless the tweeter line is properly loaded down to those freq.s.. - which is almost unheard of in a domestic speaker).

There are probably also some added problems relating to diffraction with most similar designs as well.


I guess others can hear this for themselves at a show with Legacy Audio speakers (..like their Whisper). They usually pick a large venue for them, so you can move closer to (and further away from) the speakers to hear what it's like. It's not the same as a line array, but it's not hugely different either.
 
Last edited:
Your thoughts on OB vs Sealed? With such a high Qt, I think at least aperiodic will be needed to flatten the hump in the lower midbass.

My preference is for Aperiodic for both.

Basically the tweeter should have a tapered line with a resistive "vent" at the end - bisecting or moving through the cabinet with the midbass line, terminating at the rear baffle (outside the enclosure, flush to the rear panel).

The cabinet for the midbass line should be a relatively large volume (relative to the all the driver's Vas) with a substantial "rounding" of the the edges on the vertical sides of the baffle (..i.e. closer to a "cylinder" when viewed from the front baffle). Rear resistive venting.

IF it's just a tweeter line then use the approach in the midbass line.

IF you can't do a large enclosure, then open baffle would be the alternative - but that would likely require a midbass line.
 
Last edited:
Ok so while awaiting the arrival of the Vifa's, I've been considering the midbass possibilities. Since I'll be using very capable subwoofers, I have been considering an OB approach to the midbass as well. I'm thinking about a very narrow tweeter line enclosure with a taper to point in the rear in front of a line of 6 8" midwoofer on a 12 U shape baffle with an open back. The midwoofers I have in mind are very smooth to 1.5khz bit I'm thinking a cross 1st order around 500hz on the midwoofers with 2nd order on the tweeter line.
 
So in comparing the two inch to the three inch Vifa's, with just four grouped in a simple test box without any shaping components, the 2" drivers sound more cohesive. The weather here is pretty horrible so outside measurements will have to wait a bit. In the meantime, Madisound has the Neo version on sale so I'm going to order up a few as well.
From a practicality standpoint, it's clear that I want to keep this project manageable....I mean less than a floor to ceiling array. I've been a big fan of WMTMW type systems for a while now, and can't help applying this hybrid alignment in conjunction with an array. Since I can't do any reliable measurements, I'm going to try a vertical array of 16 of the 2" Vifa's with an 8" woofer on the top and bottom...call it a WTTTTTTTTTTTTTTTTTW. I've got a sheet and a half of some nice 3/4 Auraco plywood to mock a cabinet. I'm thinking around 6ft or two meters tall with the drivers all offset 18" from the floor. I'm also thinking of the benefits of the wider baffle of 12" the entire length of the enclosure. Certainly less diffraction for the little Vifa's that will have enough of that from placement so close to each other. Also less BSC required. I'll try a target sim of 350hz second order for the cross. By the time the little Vifa's are down to 200hz, HD should be well under control, especially with 16 in service.

Now taking a page from one of PaulW's attempts with the 3" drivers, I'm considering the advantages of a slight horn loading of the array. If nothing else, it will align the drivers acoustic centers with the 8" woofers on the baffle, a strong benefit to an LR2 topology. Last night I opened a 2.25" channel in the middle of a 12" ply strip intended as a baffle. At the top and bottom of the channel, I continued the cut horizontal to within 1" of the edges. I'm gonna try and jig up a press that will push/pull the flaps inwards, creating a slight 1" depression....sort of a horn. We'll see how it goes.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.