Let's Design a Pro PA Line Array

[Randy L]

If you were to design a portable line array speaker for PA/DJ use, along the lines of the Bose PAS sticks, which drivers would you use and how would you design & build the cabinets?

I'm thinking a non-powered version able to handle a good 500 watts RMS, get nice and loud (high spl rating per driver) but very clean sounding.

Knowing a sub is needed, how would you design the thin & tall "stick" so it secures to the bass/sub cabinet? Is there some sort of twistlock insert to "snap" the stick onto the sub?

I doubt any true full range drivers would work due to the high power ratings for PA use, so I'm thinking an array of six-eight 6" (or so) extended range drivers topped with one or two quality pro super tweeters to cover the very high frequencies might be the way to go. Of course, this would necessitate the designing and building of a passive crossover network with x-over freq dependant on the selections of drivers and tweeters.

Trying the keep the stick's appearance as narrow as possible, the cabinet would be just a bit wider than the drivers used. With this in minds, how would you attach the speaker grill (I'm thinking extruded metal like PE sells here for a full faced grill)?

Other than wood, what other materials might be used? Remember that it must be well built to withstand the rigors of load ins & outs, and "gigging"

[Paul O]

You will need something that looks like this IMO... http://www.slsloudspeakers.com/content/view/264/107/
preferrably processed with DSP and powered with a seperate amp channel for each mid and high section. Subs can be reflex or horn loaded and should be positioned for best room coverage.. which most likely won't be directly under the arrays in most cases.

[Randy L]

Paul,
I've seen those designs and I've always been confused about the placement of the HF drivers, being on the side. I thought the mids might be preventing the even horizontal dispersion of the HF freq's. I've always read that the best driver alignment was vertical for the best dispersion/polar pattern, or is this old school thinking?

Last night, I too began to think that the best design for the subs might be a pair of 15" subs side-by-side, at the center of the "stage" area for the best coverage. The arrays would be standing alone to the sides. If the subs could be hidden, can you imagine the look on people's faces wondering how you're getting bass from a couple thin sticks.

[Randy L]

Forgot to add...in respect to Parts Express, let's use PE products.

Considering Paul's suggestion, what do y'all think about the Dayton planar tweeter in the array? Using 5 would achieve the 400 watts RMS goal. Maybe using 6-8 per cabinet might be better?

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=275-085

Only issue I have is that placing the planars or other HF driver beside the mids makes for a cabinet that is pretty wide...possibly 11" wide.

[Randy L]

Just got a brainstorm. How about making the cabinets deeper (maintains the thin appearance up front) to allow recessing the mids, then mount the Planars in FRONT of the mids?

[meep]

keep in mind that the bose arrays cross into the subs fairly early... it's less sub and more a "bass cab". Don't recall off-hand, it's been a few years, but I'd guess 150Hz or maybe a touch higher? 160-180?

I love the idea of this project btw...

M

[AMC]

A side note I guess, but generally when speaking of line array in pro sound, you are talking about Stack-able cabs that produce a line array, some might be confused.

One thought on a system like this might be to use a open baffle design(or at least open back), crossed over very high, maybe 400-500hz, to dual bass modules. This would reduce weight, and make it more flexible as far as drivers in the array go.

What exactly is your purpose for this? (location, size, music...)

Also, I would avoid the 'super tweeter'. they almost always sound harsh, are hard to control, and there just is not that much music up there. A good 1.? driver with a good lens would do everything you should need. Just put it at the top with a horn matched to the array width.

[Randy L]

^^ very good point about the classification of "line array"

My intent is to see if we could collectively come up with a great design for a "stick" to be used for live sound or sound reinforcement use (DJ, etc.)

Emulating the Bose PAS systems which have the numerous drivers in a vertical line array design, I think we could come up with one that looks almost as nice (slim profile & very portable), and hopefully sound exceptional and LOUD.

[wg_ski]

While you might be able to come up with something that sounds pretty good, I'd question just how *loud* you're going to get a stack of itty bitty drivers that aren't supposed to be very loud to begin with. Those Bose systems don't go very loud and you're up against the same laws of physics. How long do you think the PT2's will last? 20 seconds, IYL?

I would be thinking more along the lines of 8"/1", in something like 36" sections that could be stacked/bolted together. Two or three high for a small dance hall, go higher outdoors (but the rigging is up to you and your lawyers). Put the 8" line side by side with the HF drivers - one LF to one HF. The horn flares may need some custom machining to make it appear like one big long one that goes end to end. It may take a few iterations to get it right, but plastic horns are cheap. If the BFM cabs can do this with PIEZOs, it could certainly be done with D220's, etc.

[AMC]

While you might be able to come up with something that sounds pretty good, I'd question just how *loud* you're going to get a stack of itty bitty drivers that aren't supposed to be very loud to begin with. .

I would agree.

I think you/we should state a goal, here might be an example...

Base modules, less the 6 ft3, under 50lbs

Towers, 48 inches tall, 8 inches wide, shoot for under 50lbs.

whole system should be able to produce 130db with no more then 1000watts TOTAL power. (think 2 QSC GX3 amps)

and about 50 - 15,000 Hz range?

I'll work something out and get back to ya. Any other parameters we should set?

[AMC]

Heres an idea.

each side...
6-8 Eminence Alphalite 6A-CBMR 6" Neo Sealed Back Midrange
Mate that with a 1 inch driver and horn of suitable width.

under it, a 15 in scoop run up to 500Hz (here come the scoop haters)

The sealed back mid ranges would make the box design almost irrelevant, and it could be made very lite. The scoop should be able to deliver up to 500Hz without to much trouble, although it is far from ideal. Consider a basslite 15 for this. (C2515?) or maybe a deltalite 15. Avoid the LF drivers as their midrange tends to be less then fantastic.

The regular alphalite 6's would be great, but then the box is more important, and heavy, and they are not as efficient.


Opine away...

[billfitzmaurice]

Heres an idea.

each side...
6-8 Eminence Alphalite 6A-CBMR 6" Neo Sealed Back Midrange
Mate that with a 1 inch driver and horn of suitable width.

under it, a 15 in scoop run up to 500Hz (here come the scoop haters)

The sealed back mid ranges would make the box design almost irrelevant, and it could be made very lite. The scoop should be able to deliver up to 500Hz without to much trouble, although it is far from ideal. Consider a basslite 15 for this. (C2515?) or maybe a deltalite 15. Avoid the LF drivers as their midrange tends to be less then fantastic.

The regular alphalite 6's would be great, but then the box is more important, and heavy, and they are not as efficient.


Opine away...Crossing from the sub to the top at 500 Hz is poor practice, as is using scoops, which are reviled for good reason. And placing the subs under the tops is wrong in so many ways. Stick with standard Alpha/Beta 6s or 8s for a 100 to 150 hz crossover and cluster the subs. There's also no advantage to 1" throat compression drivers or planars, KISS with a vertical array of piezos.

[Paul O]

I'm old enough to remember when scoops were cool and very popular, but there are so many much more capable designs now without the compromises.. so why bother.

Back to the project..

It seems the column speaker is making a resurgence, several big name companies are now marketing new install products based on the idea including Tannoy Qflex and the JBL CBT series. This JBL design puts the tweeters in a row in front of the mid column..
http://www.jblpro.com/catalog/general/ProductFamily.aspx?FId=89&MId=2

[Randy L]

Bill,

I'm tending to agree with you that the 500hz rolloff of those Alpha-6's may prove to be a bit too high. On your suggestion of the piezo's, I've used them quite a bit in various designs and feel they sound a bit too harsh. Is there any type of passive filtering to smooth them out? I've even read that some have lightly stuffed the mouths of the piezo's with some mild fabric or foam to tame them.

On the above remarks about the itty-bitty drivers...YES...I agree. I just didn't want to limit the imagination by saying the drivers had to be minimum size "X".

My thinking is a grouping of 8 quality pro driver in the 6" size, crossed over at around 150-200hz to the bass bins. I kinda like the idea of leaving the "sticks" alone to each side of the stage (not rock concert levels, but more a small to meduim sized venue) and clustering the bass cabs...hidden if possible.

[Paul O]

Emulating the Bose PAS systems which have the numerous drivers in a vertical line array design, I think we could come up with one that looks almost as nice (slim profile & very portable), and hopefully sound exceptional and LOUD.

There's a project in the showcase http://www.parts-express.com//projectshowcase/indexn.cfm?project=Kuze3201 that's almost exactly what you describe but it won't go very loud and just like the PAS needs an additional bass module.



How long do you think the PT2's will last? 20 seconds, IYL? Are they really that fragile? How robust would a stack of 8 be wired series/parallel for 4ohms? I was considering those for a project like this but wondered if they could produce the output to match 8 Dayton PA-165's.

[Randy L]

I'm quite intrigued by the Dayton Planar tweeter and wonder how it might actually sound in a PA application.
http://www.parts-express.com/pe/pshowdetl.cfm?&DID=7&Partnumber=275-085&ctab=2#Tabs

Taking cues from the JBL's CBT line, what about placing 7 of them in FRONT of 8 mid drivers? The planars are 3-1/2 wide so they wouldn't hide/block too much of the mids. Otherwise, like Bill suggested, maybe a stack of piezo's in front of the mids? While piezo's (by design) don't need crossovers, I found they do sound better when crossed over at around 2.5khz.

[Paul O]

I'm quite intrigued by the Dayton Planar tweeter and wonder how it might actually sound in a PA application.
http://www.parts-express.com/pe/pshowdetl.cfm?&DID=7&Partnumber=275-085&ctab=2#Tabs. Yeah me too.


Otherwise, like Bill suggested, maybe a stack of piezo's in front of the mids? While piezo's (by design) don't need crossovers, I found they do sound better when crossed over at around 2.5khz.

And there's another rub... one of the advantages of a proper line array is wide polar response, but to maintain that over the whole spectrum crossover frequencies have to be below the point where drivers begin to beam excessively, and that means a 2khz crossover for the tweets at the most.

[Randy L]

What about this Selenium 6" woofer?
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-422
I like the RMS rating, though I'm not too confident in the 9khz response. According to the freq. response curve, it looks like it becomes "unstable" above 2khz. Cost ain't too bad neither.

[billfitzmaurice]

On your suggestion of the piezo's, I've used them quite a bit in various designs and feel they sound a bit too harsh.Then you haven't used them in a line array. They aren't very flat, but it's pro-sound, where DSP isn't an option, it's a necessity, so flat response is moot.

[wg_ski]

PT2's...


Are they really that fragile? How robust would a stack of 8 be wired series/parallel for 4ohms? I was considering those for a project like this but wondered if they could produce the output to match 8 Dayton PA-165's.

They're not especially fragile - any more than any other home stereo tweeter. Which experience has shown time and again can't be used for any serious PA app. People get the idea that a "tweeter" will do just as well because it can handle the same 40 watts as a standard 1" compression driver. But at 10dB less sensitivity, you ask it to do 10X the work, and that's where you run into trouble. I know of the argument that "there isn't much up there" because of the typical power distribution of full range music. But you can't rely on that where a speaker may be subject to full power peaks in the vocal range, or feedback, or clipping....

[wg_ski]

Then you haven't used them in a line array. They aren't very flat, but it's pro-sound, where DSP isn't an option, it's a necessity, so flat response is moot.

Can you even still GET real ones anymore? Like really equivalent to the original Motorolas? Without spending the same money you would on D220's?Everything I've seen lately have been the $1.95 apiece Asian copies which truly do sound like garbage DSP or no DSP. The biggest problem with the cheap junk is they're about 85dB/W instead of the 94 they used to be. No DSP will correct that, you just have to drive them hardrer. And when you put 10x the power into them they do stress and sound stressed.

[Randy L]

Then you haven't used them in a line array. They aren't very flat, but it's pro-sound, where DSP isn't an option, it's a necessity, so flat response is moot.

Bill,
you are correct. I have yet to design or build a line array speaker. I have clusterd four of the "Radio Shack" piezo's on a 3 way PA speaker though.
On those piezo's what brand and model do you recommend, and how many?

[AMC]

Crossing from the sub to the top at 500 Hz is poor practice, as is using scoops, which are reviled for good reason. And placing the subs under the tops is wrong in so many ways. Stick with standard Alpha/Beta 6s or 8s for a 100 to 150 hz crossover and cluster the subs. There's also no advantage to 1" throat compression drivers or planars, KISS with a vertical array of piezos.

I did warn of the scoop hate...

You are correct about crossing a sub at 500Hz, but it's not a sub, it's a scoop. And whats the hatred for scoops? (rhetorical question, their flaw is well understood, and hard to hear) Secondly, I would hardly call a vertical array of piezos 'keeping it simple', for someone who does that all the time, maybe, otherwise... not so much. A good single compression horn would fit rather nicely, and as stated this is supposed to be very portable, carting around fewer electronics would be beneficial (less signal processing). My only concern with the regular alpha would be their (stated) lack of sensitivity. Not sure they would get loud enough for the requested application. That being said, this type of design will come with trade offs, Maybe that's one of them.

[wg_ski]

A scoop crossed over at 500 Hz would not be a sub - it would be a "bass bin" or a even a "low mid". Crossing that high you won't have much of a "line array" benefit to the lines - you may as well go with something simpler. Like a single 10"/1" with neo drivers.

A small line of Alpha 6's isn't going to go all that blasted loud. Neither will a 15 inch scoop (before overexcursion). If you're just trying to compete with Bose, you probabaly could on price -their drivers are likely just $10-specials. Loudness I don't know - unless you did build 8 foot lines I wouldn't think there would be much benefit. And of course I'd use a real subwoofer, even if it was just a 15" manifold that just hits 45 Hz.

[AMC]

- you may as well go with something simpler. Like a single 10"/1" with neo drivers.



I think the original point was more for aesthetics then performance. Don't think of it as a line array, think tall very thin pro cabs, that is the design goal. For moderate sized venues, I think it is a fully attainable goal.

[AMC]

Then you haven't used them in a line array. They aren't very flat, but it's pro-sound, where DSP isn't an option, it's a necessity, so flat response is moot.

I understand your point, but with this logic, why worry about making your speakers sound good at all? just make them loud and let the DSP correct all their problems. Seems a better goal is to make the speakers sound good, and let the DSP correct for room and environment problems.

just $0.02

[billfitzmaurice]

Can you even still GET real ones anymore? Like really equivalent to the original Motorolas? Without spending the same money you would on D220's?Everything I've seen lately have been the $1.95 apiece Asian copies which truly do sound like garbage DSP or no DSP. The biggest problem with the cheap junk is they're about 85dB/W instead of the 94 they used to be. No DSP will correct that, you just have to drive them hardrer. And when you put 10x the power into them they do stress and sound stressed.
I've tested Goldwoods and original Motorola and the later CTS side by side. The Goldwoods are better.
When there was nothing but Motorola the gear snobs looked down on them because they were Motorola, not JBL or EV. When it passed on to CTS the gear snobs said the Motorola's were better, even though they were identical. When CTS went away some outfits sold NOS for ten times the price they originally went for, and then the gear snobs said that the NOS Motorola and CTS were fine but the Asian stuff wasn't.

why worry about making your speakers sound good at all? just make them loud and let the DSP correct all their problems.Because it puts you in good company? No top of the line pro-touring rigs can even be used without DSP. I believe that one shouldn't have to pay $10k per cab for the same privilege.

[Paul O]

I had a bunch of the original Motorola KSN1005 piezos and they were quite good when used for what they were designed for... as super tweeters. But when I finally blew them the replacements never matched very well and that has pretty much been the story ever since, if you want 2 that sound the same you probably have to buy 6 or 8 and pick the two that are closest. I don't know.. haven't been a fan of these since than so they wouldn't be my first choice but I'm sure you could make use of them if you really wanted to.


As for the design intent of this project, there are now lots of guys using the Bose PAS for DJ apps and it seems to be sufficient for smaller gigs where big SPLs aren't needed.. think cocktail parties and other adult events. It certainly is portable but the bass modules are a bit of a joke and I don't like the way the controls are built into the base of the array.. I'd rather it was all rack mounted and the speakers passive.. that would open up some mounting possibilities for the sticks to get them up off the floor.
I don't think it would be much challenge to exceed the performance of this system but doing so in as small a package is going to be the challenge.

[Randy L]

Agreed Paul. Many fellow DJ's on another site do state that for small, and even some medium sized venues, using two Bose PAS sticks and 2 bass modules is very satisfying. But, they admit that when the crowd got really partying and cried for more volume or bass (booty songs), the Bose couldn't keep up.

These "sticks" would be well thought out and capable of enough (clean)volume to be used in large venues or school dances. I'd even like them to be useable in smaller venues for a live band.

While I am comparing these to the Bose PAS (in design elements only), my main target is to allow these to replace the common 15" full range PA speakers used in the industry.

Thus, my thoughts on also designing a pair of woofer cabs using 15" drivers....I'm seriously interested in trying out the "new" Eminence Kappalite Neo woofer.

Brainstorming the "sticks" I think building a 2 piece section that twistlocks together, but looks like a single column. The bottom would be a hollow base to elevate the top "half" that contains the drivers, with the top driver at about 76" high.

Now I just thought of something. How would this speaker sound on a typical stage (about 3 feet high?), the vertical line array would be way over the crowd's heads.

[Paul O]

Brainstorming the "sticks" I think building a 2 piece section that twistlocks together, but looks like a single column. The bottom would be a hollow base to elevate the top "half" that contains the drivers, with the top driver at about 76" high. I'd just use a regular speaker stand, the base would look cleaner but it's not as stable and a stand. I guess having the option is a good idea but If I was building 4 boxes I'd want them all filled with drivers. :D If the "array" is 4ft high for example ideal placement for DJ events would be about 3ft off the floor and perpendicular to the floor, that will cover everybody in a seated and standing positions. One of the things PAS users comment on is that a lot of the sound is blocked when people crowd around the speakers.. that's gonna happen when the array is only 6ft high and standing at floor level. And IMO all those drivers below knee level are wasted, so using a shorter array with an adjustable stand makes more sense.. though that shorter array will have to use drivers capable of more output.


Now I just thought of something. How would this speaker sound on a typical stage (about 3 feet high?), the vertical line array would be way over the crowd's heads. For stage use the speaker will have to be angled down somewhat I think which is one of the alternate mounting scenarios I was referring to above, if the box has flypoints it could be hung from a truss and angled as required. This is a an application where stacking 2 boxes would be beneficial, you get added output as well as much greater coverage.

[Randy L]

Good points Paul! :cool:

[AMC]

I know years ago, the general thought was that you would want the speakers aiming above the heads of the people in front, this way they compensate for their proximity with the reduced response being off-axis. Like everything else in the audio world, I'm sure this has changed.

As far as a driver for the arrays, what about LA8-CNMB, lite being a neo driver, about 95db/watt, capable of about 117db each, should be able to get about 120 each side at 250watts, driving them at about 1/4 power for a good long life. and they go down to about 125Hz at that SPL. Should only need 4 per side, keeping it at 8 ohms, and they are truncated on 4 sides for close placement and a narrow width. And they go up to about 3000Hz, easy for a 1" CD to take over from there.

Eminence Kappalite 3015LF Neo 15" Driver X 1

Eminence LA8-CNMBA Neo 8" Midrange X4

Eminence ASD1001S 1" HF Titanium Horn Driver 1-3/8"-18 TPI X1
Selenium HM17-25 1" Bi-Radial Horn 60x40 1-3/8"-18 TPI X1

This would be per side of course. All PE, all Eminence, and all but the CD are NEO drivers

Does get spendy though (1,385.62 just for drivers)

Should achieve about 120db, with no more the 250w per side (passive) or 500w per side active X-over.

The arrays should be 9 inches wide, and about 40 inches tall.

I would guess, this should do a Mid sized venue (+/- 150) with music, or a larger (+/- 300) for light music, PA duties.

[Paul O]

As far as a driver for the arrays, what about LA8-CNMB, And they go up to about 3000Hz, easy for a 1" CD to take over from there.

OK.. but there are a few problems with that. It takes a lot more than 4 drivers to produce line array, and even if those did begin to behave as an array the tweeter is a point source so amplitude response will be mismatched everywhere except the distance where it is calibrated to be the same. And 3khz is way to high a crossover to maintain a somewhat uniform polar response, for an 8" driver that has to be below 1.7khz. Even using 6" mids and tweeters that can support a 2khz crossover is a bit of compromise as far as polar response is concerned but it may be "close enough" with music to deliver what appars to be that ultra wide dispersion line arrays are known for.

Designing a speaker system that sounds good everywhere on a 50ft wide dance floor is a lot different that building one to sound good in a single listening position. A true line array has a distinct advantage over point source speakers, but the reality is you need an 11ft high stack-o-speakers to deliver true line array properties down to 100hz so there are going to be compromizes made.

[Randy L]

On the issue of a crossover point, what about finding a driver that is pretty smooth up to 6000hz and set the crossover around there? Also, what slope is best for a line array...6db/12db/etc?

An example is this Tang Band woofer.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-850

Maybe 20 drivers per cab, and at $22.45 per driver, the cost is fairly good. At a safe 40 watts per driver allotment, that's 800 watts RMS that it could handle. At 88db spl, what would the spl be for the 20 drivers together? Cross it over at 150hz to the outboard woofer. The response curve seems like it could handle a 6khz crossover.

[Paul O]

On the issue of a crossover point, what about finding a driver that is pretty smooth up to 6000hz and set the crossover around there? That would be fine as long as the driver are 2" in diameter or less.


Also, what slope is best for a line array...6db/12db/etc? The crossover slope has nothing to do with the speaker type, what works best will be determined by the response of the individual drivers just like it is with a simply 2-way.

[Randy L]

That would be fine as long as the driver are 2" in diameter or less.

I don't think a 2" driver would be found to withstand the hard duty of a PA application.

On my example Tang Band woofer, I just noticed it's a little over 6" in outside diameter. In a standard vertical alignment, 20 drivers would make this a 10' tall cabinet!!!!! Whoa! Too tall! :eek: Maybe 12 or 14 drivers bringing the cabinet to 7 feet tall.

Update: I inputted the T/S specs of that Tang Band woofer into Win ISD Beta and with each driver in its own .167 c.f. sealed enclosure, I inserted 12 drivers as the number. Win ISD shows a sensitivity (1W/1M) of 88db @ 150hz. It then goes up dramatically, peaking at 105db @320hz, then slopes back downward. Win ISD only lists up to 1khz, and the graph shows that the 12 drivers will have an SPL of 99.5db.
In the option to set the power to the "project", I'm not sure if I'm supposed to put the 40 watts I'm allotting per driver, or rated wattage of an amp. In my case, I have a 450 watt/ch amp. So, with the 12 drivers, at the 40 watts, WinISD shows a peak SPL of 121.5sb SPL @ 323.94hz, and 115.5db SPL @ 1khz. When I plug in the 450 watts my amp can push to the drivers, it shows 131.5db SPL @322hz, and 125db SPL @1khz.

[AMC]

In the option to set the power to the "project", I'm not sure if I'm supposed to put the 40 watts I'm allotting per driver, or rated wattage of an amp. In my case, I have a 450 watt/ch amp. So, with the 12 drivers, at the 40 watts, WinISD shows a peak SPL of 121.5sb SPL @ 323.94hz, and 115.5db SPL @ 1khz. When I plug in the 450 watts my amp can push to the drivers, it shows 131.5db SPL @322hz, and 125db SPL @1khz.

That might be fine, but they have an RMS limit of 28watts. Stick with pro sound gear for pro sound applications.

[AMC]

OK.. but there are a few problems with that. It takes a lot more than 4 drivers to produce line array,

We know, it's not really a line array, as discussed before, just a tall thin speaker.

[Paul O]

I don't think a 2" driver would be found to withstand the hard duty of a PA application. What's hard about it? A 2" driver is small and lightweight so it not as easily damaged from being bumped around, and in an array will only have to handle 15w or so. That is part of the beauty of an array, you don't need big hefty high powered drivers when the load is shared over a whole lot of them.

[billfitzmaurice]

We know, it's not really a line array, as discussed before, just a tall thin speaker.
Any speaker with two or more elements vertically aligned is a line array. Forget about all the white paper piffle and formulaic mumbo jumbo, concentrate on what matters: does it sound good?
At some frequency and distance every speaker is a line source and every speaker is a point source. Caterwallering over whether it's a line array or a line source or a tall thin speaker is purely philosophical masturbation. Figure out a design goal and achieve it already.

[SpasticColin]

Any speaker with two or more elements vertically aligned is a line array. Forget about all the white paper piffle and formulaic mumbo jumbo, concentrate on what matters: does it sound good?
At some frequency and distance every speaker is a line source and every speaker is a point source. Caterwallering over whether it's a line array or a line source or a tall thin speaker is purely philosophical masturbation. Figure out a design goal and achieve it already.

Parole Board chairman: You're not just telling us what we want to hear?

H.I.: No, sir, no way.

Parole Board member: 'Cause we just want to hear the truth.

H.I.: Well, then I guess I am telling you what you want to hear.

Parole Board chairman: Boy, didn't we just tell you not to do that?

H.I.: Yes, sir.

Parole Board chairman: Okay, then.

-- Raising Arizona

[AMC]

What's hard about it? A 2" driver is small and lightweight so it not as easily damaged from being bumped around, and in an array will only have to handle 15w or so. That is part of the beauty of an array, you don't need big hefty high powered drivers when the load is shared over a whole lot of them.

A. I think it would be hard to come up with 2" drivers that would be efficient enough, even with the use of several.

2. You would still have the problem of not going low enough, maybe 300HZ at best in a high SPL application.

To be honest, as much as some people hate Bose, they build a decent product here, and to build anything better is a challenge. That is a hard system to beat when you look at it's size and weight.

Still like the 8" neo drivers, One problem I would see in NOT using neo drivers, if you have this tall thin speaker, and it starts to get heavy, and it will, it will become very unstable.

As far as construction goes, I was thinking about a triangular-ish design, where it would look a bit like the cross section of a boat hull when viewed from the top. You could construct it like a boat, triangular shaped ribs between each driver, and triangular end caps, then use bendable ply (two layers min) on the sides. The triangular shape would cut down on you volume, but it would look cool.

[Randy L]

AMC,

I agree that the Bose PAS does sound very nice indeed. I've heard them and know a handful of DJs that use them. For professional appearance, they can't be beat. For small to medium sized venues where music doen't need to be loud they sound incredible. But...they crap out when bass heavy (booty) songs are played and when the crowd gets partying and want it louder.
Thus, they wouldn't be useful for a school dance, rock club, or large hall for a corporate event.

Yes, weight should be an important factor. Don't want a 60lb speaker! But, out of lack of available plastic or polycarbonate materials, and the tools to work with them, almost all us us will have to resort to wood which is inherantly heavier.

I was thinking the same thing about a extreme trapezoidal design too.

[Music is life]

I would decide what output would be the max needed.
Decide how high you would go, that being the limit.
Build half of that height and make it stack-able maximum 2 unites.

If woofers and tweeters, don't go all the way down to the lowest adjacent woofer but do go all the way to the top of the cab with both woofers and tweeters. With mirror image cab stack that one up side down on top when 2 are needed.

If tiny little drivers like the Bose were used try cross firing.
If you are just trying to beat the Bose use this:
http://www.parts-express.com//pe/showdetl.cfm?&Partnumber=295-010
and this:
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=275-025
or piezos.

For higher spl, 4 alpha 6a's and a lower crossing tweeter. Or 2 of these near the top of the cab
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=294-600
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=294-618
next to the woofers, crossed at 2400Hz.
Alpha's crossed at 120Hz (watch that mechanical limit, only 4mm). That's about 124db long term (4 alpha 6a's). About 130db for 2cabs.
Will be trying the latter myself very shortly since I have all the parts.

Just some ideas.

[Paul O]

A. I think it would be hard to come up with 2" drivers that would be efficient enough, even with the use of several..

Several? With drivers that small it's going to take a 6' high stack to produce reasonable output, but every time you double up the number of drivers sensitivity goes up so starting with 84db@1w drivers really isn't an issue.


2. You would still have the problem of not going low enough, maybe 300HZ at best in a high SPL application. Yes that will be a problem, the Bose "bass module" goes all the way up to 250hz or so as a result of the fact they use 2.25" drivers.

I'm not saying this is the best solution here just trying to clear up some misconceptions. This can be done with really small drivers it just takes a lot more of them, but again there are compromises to consider.

There are no tweeters in that design so no crossover either but while there is decent highend on axis the horizontal coverage is actually worse than a simple 2-way PA box with a 1" CD horn above 6khz, so much for Bose's claim of 180 deg coverage. And the on-axis response isn't that great either.. the Panarray MA12 which is the install version of the PAS is rated -3db at 12khz for example. How's that go again.. no highs.. no lows... must be...

So.. I still prefer a 2-way design with larger mid drivers and smaller tweeters, using 4" to 6" mids and either planar or small dome tweeters will cover 100hz-20khz without much problem and deliver line array properties over most of that range as long as the column is at least 4ft high. The challenge will be keeping the weight down so neo drivers would certainly be better, and I don't think the cabs would weigh that much if say 9mm birch ply was used to build a trapazoid enclosure with routed joints and integrated bracing. After all the box doesn't need to support heavy bass just house the drivers and be resonant free.

[AMC]

Several? With drivers that small it's going to take a 6' high stack to produce reasonable output, but every time you double up the number of drivers sensitivity goes up so starting with 84db@1w drivers really isn't an issue.



But to get to 120db, at 5 watts a driver, you would need somewhere around 200. Not exactly practical. (and that makes 1000 watts a side)

[Paul O]

But to get to 120db, at 5 watts a driver, you would need somewhere around 200. Not exactly practical. (and that makes 1000 watts a side)
Your math is a bit off... with about 500w into a column of 32 drivers each one will see about 15w and output will easily exceed 120db at 1m.

P.S. As stated earlier every time the number of drivers is doubled there is a 3db increase in sensitivity. 32 drivers is 15db more sensitive than one(1 to 2 to 4 to 8 to 16 to 32 or doubled 5 times) so if that driver has a 1w rating of 84db for example the array will have a 1w rating of 99db.

And 500w is 27db more than 1w so the potenitial max output is 99db + 27db = 126db.

[AMC]

Your math is a bit off... with about 500w into a column of 32 drivers each one will see about 15w and output will easily exceed 120db at 1m.

P.S. As stated earlier every time the number of drivers is doubled there is a 3db increase in sensitivity. 32 drivers is 15db more sensitive than one(1 to 2 to 4 to 8 to 16 to 32 or doubled 5 times) so if that driver has a 1w rating of 84db for example the array will have a 1w rating of 99db.

And 500w is 27db more than 1w so the potenitial max output is 99db + 27db = 126db.

First off, show me a 2" driver that can handle 15watts in a pro sound application.

Second, 99db 32 drivers, 1 watt, = 111db at 16watts each.
to get to 120 you would then need... 256.

When you add a drivers: 1 driver 84 db @ 1 watt, add another and you get 87db with both driven at one watt, not with one watt divided between the two of them. so In your example of 99 db per watt, that is with one watt each, not one watt total.

[Paul O]

First off, show me a 2" driver that can handle 15watts in a pro sound application. . Doesn't matter what the application is, power is power.



When you add a drivers: 1 driver 84 db @ 1 watt, add another and you get 87db with both driven at one watt, not with one watt divided between the two of them. No you get 3db from doubling the cone area and another 3db if you also double the power applied for a total of 6db gain.

[AMC]

Doesn't matter what the application is, power is power.


No you get 3db from doubling the cone area and another 3db if you also double the power applied for a total of 6db gain.

Then you show us Paul, Build it, run it at 120db for a few hours without a break, and prove your the man. I'll be here waiting.

(I predict lots of smoke and distortion, not necessarily in that order.)

[Randy L]

Second, 99db 32 drivers, 1 watt, = 111db at 16watts each.
to get to 120 you would then need... 256.


Don't mfg's rate the peak SPL's at the speaker's (or driver's) peak power rating? So the 120db target (maybe even higher) would be calculated using the individual driver's power rating. For a margin of safety from blowing drivers, I would only power each driver at its RMS rating, or maybe a bit higher...but never its peak rating.

[Paul O]

Don't mfg's rate the peak SPL's at the speaker's (or driver's) peak power rating? So the 120db target (maybe even higher) would be calculated using the individual driver's power rating. For a margin of safety from blowing drivers, I would only power each driver at its RMS rating, or maybe a bit higher...but never its peak rating.

Speaker power ratings are subject to a lot of conditions just like any other factor. The difference between the rms and peak ratings is an indication of what type of signal the manufacturer is using to test the speakers but without knowing how long the test was for or what spectrum the signal consisted of it's hard to draw accurate conclusions. The two common signals used are pink noise and sine wave tones, pink noise will often have a 6db crest factory while the sinewave is 0db or a steady state tone. If there is only 3db difference between the RMS and Peak ratings then some other test signal was used or it was done with short duration burst like maybe a 1ms sinewave burst for example because a steady state tone will heat the voice coil much quicker due to it's higher average power level. All this makes comparing the ratings of different drivers difficult and really doesn't tell you much about a drivers music power handling since it can have crest factors of 10-15db or more, but in general it's not uncommon to discover a driver will handle quite a bit more than it's "rated" power with dynamic music as long as it doesn't exceed driver excursion limitations.
It is generally accepted that the RMS rating is the thermal power handling capacity of the voice coil when driven with a sine wave for some period of time, but again if this additional info isn't specified it's hard to know. Sometimes manufacturers will state how thier drivers are tested in additional documentation available on thier web site.

[Paul O]

Then you show us Paul, Build it, run it at 120db for a few hours without a break, and prove your the man. I'll be here waiting.

(I predict lots of smoke and distortion, not necessarily in that order.)

:confused: It's clear you don't understand as much about speakers as you think you do. There is lots of data available "out there" that will verify what I have presented, but previous experience has taught me you can lead a horse to water but you can't make him drink so I'll leave it to you to go find it.

[Music is life]

1 = 84
2 = 87
4 = 90
8 = 93
16 = 96
32 = 99

99 = 1
102 = 2
105 = 4
108 = 8
111 = 16
114 = 32
117 = 64
120 = 128

128 / 32 = 4

[AMC]

:confused: It's clear you don't understand as much about speakers as you think you do. There is lots of data available "out there" that will verify what I have presented, but previous experience has taught me you can lead a horse to water but you can't make him drink so I'll leave it to you to go find it.

Then build it and show me Paul, like I said, I'll be here waiting....

[wg_ski]

Sometimes manufacturers will state how thier drivers are tested in additional documentation available on thier web site.

That's true if you're building with Eminence, JBL, EV, or even Selenium on the 'pro' side. Or one of the regulars from the Tymphany group (Scan, Vifa, Peerless...) you may find actual test data. But if you're going after a bunch of small 'budget' drivers your guess is probably as good as anyone else's for power handling. For small midwoofers, ROT is 40 watts per inch diameter squared, lacking other info. Kapton will buy you up to 2x.

[Paul O]

Then build it and show me Paul, like I said, I'll be here waiting....

I don't see the point.. would you believe my results if I did? I got nothing else to say on this.. if you don't believe it's possible then I can't help you.

[AMC]

I don't see the point.. would you believe my results if I did? I got nothing else to say on this.. if you don't believe it's possible then I can't help you.

What you are talking about is acoustic voodoo. They idea of getting something for nothing never works. By your idea, you could power an entire arena with a 100W amp if you just put enough speakers on it. What a waste for all the groups touring out there to have purchased thousands of watts of amplification, when all they really needed was a stack of 500 2" speakers. If you have sources that prove me wrong, show me. It's a fairly simple rule..

2X surface area, +3db, but only if the power stays the same, if you halve the wattage to each driver, the output from each drops 3db, canceling out the increase in number of drivers, thus +0.

with our 87db driver, 1 driver @ 1 watt = 87 db
if you add a driver, and each gets 0.5 watts, each puts out 84db, together they produce 87db

There is no such thing as a free lunch.

[SpasticColin]

Your frequency-dependent and HF directivity-reduced lunch is served. (http://www.rane.com/par-c.html#coupling)

[AMC]

This is a very entertaining site, not sure it address the question directly, but I still like it. Thanks.

[Music is life]

2X surface area, +3db, but only if the power stays the same, if you halve the wattage to each driver, the output from each drops 3db, canceling out the increase in number of drivers, thus +0.

with our 87db driver, 1 driver @ 1 watt = 87 db
if you add a driver, and each gets 0.5 watts, each puts out 84db, together they produce 87db

There is no such thing as a free lunch.

Let's see if I can follow your point.
We build a full range speaker that uses 2 delta 12lf (94db one watt) for lows and 2 beta 8a (94 db one watt) for mids and 1 apt 80 for highs. The woofers and mids are run in parallel.
Are we attenuating the tweeter about 4 to 5 db or about 7 to 8db?

[AMC]

Let's see if I can follow your point.
We build a full range speaker that uses 2 delta 12lf (94db one watt) for lows and 2 beta 8a (94 db one watt) for mids and 1 apt 80 for highs. The woofers and mids are run in parallel.
Are we attenuating the tweeter about 4 to 5 db or about 7 to 8db?

First thought, attenuate it about 40 - 50 db, (my ears don't like super tweeters, nothing against them, they just don't agree with me, kind of like baby octopus sushi)

That is an interesting, but loaded question. The apt 80 has a range at about 105db/watt, and a range at about 100db/watt. If you assume (yea, I said it) 105, then I would consider coming down about 7-8. I also tend to assume that rated efficiency on cone drivers is about 1-3 db over rated. NOW, all that being said, all this theory comes down to just that. If you could plug it all in a computer and have it spit out an accurate design, that would be great, and you are done, but real world never works out like that. That's why you have to build it and see.

This is where, as I understand him, Paul's theory would suggest you would not need to attenuate, 94db, +6db for adding the second driver, 100db, +3db for having them in parallel and thus increasing your power = 103db. Maybe I am misunderstanding him though.

I'm guessing that you have already done so, and your solution was....

[Paul O]

This is where, as I understand him, Paul's theory would suggest you would not need to attenuate, 94db, +6db for adding the second driver, 100db, +3db for having them in parallel and thus increasing your power = 103db. Maybe I am misunderstanding him though.

Yeah you misunderstand, like I thought presenting 3rd party evidence to you does no good, you obviously have not read the Rane link or researched any other information on "mutual coupling". If you did you would find it's not my theory at all.

Two low frequency drivers placed close together will produce a 3db gain in sensitivity.. that's it, but precisely what SPL that happens to be will depend upon their response at crossover frequency when combined with the crossover and any compensation elements(BSC, Zobel, etc) you happen to employ.

[Music is life]

First thought, attenuate it about 40 - 50 db, (my ears don't like super tweeters, nothing against them, they just don't agree with me, kind of like baby octopus sushi)

That is an interesting, but loaded question. The apt 80 has a range at about 105db/watt, and a range at about 100db/watt. If you assume (yea, I said it) 105, then I would consider coming down about 7-8. I also tend to assume that rated efficiency on cone drivers is about 1-3 db over rated. NOW, all that being said, all this theory comes down to just that. If you could plug it all in a computer and have it spit out an accurate design, that would be great, and you are done, but real world never works out like that. That's why you have to build it and see.

This is where, as I understand him, Paul's theory would suggest you would not need to attenuate, 94db, +6db for adding the second driver, 100db, +3db for having them in parallel and thus increasing your power = 103db. Maybe I am misunderstanding him though.

I'm guessing that you have already done so, and your solution was....

Yes, already done it and I like it a lot.
The apt 80 is 105db. It does go down a bit but well above crossover (3500hz).
Now I'm not trying to prove anyone right or wrong just telling you my experience.

Found the rise in the responce of the beta 8's less when 2 are used together. Crossed them over at 230hz third order.

Would build them again in a heartbeat but with the 3012lf's and a B&C tweeter and cross over a little lower.

I attenuated the apt about 4 or 5 db and it matched the rest of the system according to my ears.
Attributed that to the coupling effects of using two drivers which is supposed to increase the sensitivity by about 3db. Then also they were at 4 ohms so then another 3db. That's 100 to 101 against the 105 of the apt. So 4 to 5 db attenuation.

They sounded good to me.
Not saying this would work forever as other limits come into effect. Or that it works equally at all frequencies.

There may be something to learn from this "experiment":
http://www.parts-express.com/projectshowcase/Kuze3201/Kuze3201.html

And that might be relevant to the project in question.

[AMC]

There may be something to learn from this "experiment":
http://www.parts-express.com/projectshowcase/Kuze3201/Kuze3201.html

And that might be relevant to the project in question.

This was an interesting project, and pretty much exactly what were talking about. I'd be interested to know more about the final outcome of it.

Your Beta 8a's are a great example of why we should look at the response graph instead of just going by the rated number. It has a range where it runs about 94 (100 - 900Hz), and then it has a range where it runs about 97 (1100 - 4000Hz) currently rated at 95db. Depending on what your after, that's a big difference.

Lies, Damn Lies, and Specifications.

[Music is life]

This was an interesting project, and pretty much exactly what were talking about. I'd be interested to know more about the final outcome of it.

Your Beta 8a's are a great example of why we should look at the response graph instead of just going by the rated number. It has a range where it runs about 94 (100 - 900Hz), and then it has a range where it runs about 97 (1100 - 4000Hz) currently rated at 95db. Depending on what your after, that's a big difference.

Lies, Damn Lies, and Specifications.

Yea, even 2" drivers suffer from comb filtering when used as tweeters.
No doubt the same thing happens with the Bose and is compensated for with processing.
I would definitely prefer using separate woofer and tweeter. Small dome's or piezos for tweeters economies sake. Much better match for the woofer and more spl.
Here is an idea for this project:

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=290-400

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-1002

That would be 4 alpha 6's (wired to 8 ohms) to 18 Vifa (wired to 8 ohms. Two 4 ohm groups of 9 wired is series). (Then unites of the above for vertical stacking.) If my math is right they should be at the same output sensitivity and be around 99db with one watt.
But my math could be wrong. Crossover around 1800Hz.

[craigspieker]

My application is not finished and not hi end, but it's a line array with 4" drivers, tweeters to be added soon. made from sonotube with a front baffle clamped inside. Sure it probably resonates, I'll test it one of these days. But it's super light and is in two sections that latch together.

No one has mentioned the advantage of true line arrays in a large space, SPL does not fall off at the same rate at distance, more even crowd coverage.

[Randy L]

Craig,
That is true about the fall off. In the pic, I see the bottom section using different drivers. Is there a specific reason for that? Also, what kind of latch did you use?

I see a guitar...is that for size reference or do you play it through that LA?

[craigspieker]

Craig,
That is true about the fall off. In the pic, I see the bottom section using different drivers. Is there a specific reason for that? Also, what kind of latch did you use?

I see a guitar...is that for size reference or do you play it through that LA?

Same drivers, just repurposed from a project requiring white! Not a perfectionist project on this one. Latches are like you'd find on a big trunk, I just went to the hardware store.

the guitar is for size reference but yes I play acoustic and electric and sing through this set up. Electric guitar is great through this thing, Makes a loud cheap guitar amp if you have an amp emulator. Vocals not perfect running the 4" full range, need to put on some tweeters (I have some dirt cheap plastic domes on a 4" spider I'll be mounting), but considering the whole LA is probably sub $40 i'm pleased.

I've also thought about a matching sub using sonotube, mounting driver in a sealed enclosure and then latching a second chamber on bottom of it with a port to make a huge yet portable bandpass box.

[jeff_free69]

This is a really nice tweeter you can put at the top of your array if you've got mids that can go up to at least 5k.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-360.

Don't bother with any hifi tweeters - ain't worth the risk or cutting all them holes.
-------

Just to throw in my 2 cents on this whole project... I've been contemplating the same for a couple of years. Its a sexy idea but I still haven't been convinced an appropriate pro quality driver really exists for a true "stick" . I played in a band that uses the Bose PAS and the thing everybody forgets is that the "PA" in a the Bose PAS stands for "Personal Amplifier". 1 or 2 do not have the cajones to replace a true PA. By the time you get enough of them plus the additional bass units , you're in for at least 5 grand. And you still end carrying a load of gear, although no single piece is too large or heavy

Of course I'm still hoping we find that holy grail..

[AMC]

Just to throw in my 2 cents on this whole project... I've been contemplating the same for a couple of years. Its a sexy idea but I still haven't been convinced an appropriate pro quality driver really exists for a true "stick" . I played in a band that uses the Bose PAS and the thing everybody forgets is that the "PA" in a the Bose PAS stands for "Personal Amplifier". 1 or 2 do not have the cajones to replace a true PA. By the time you get enough of them plus the additional bass units , you're in for at least 5 grand. And you still end carrying a load of gear, although no single piece is too large or heavy

Of course I'm still hoping we find that holy grail..

I think your correct here, it seems to me that even if you go up to 6" drivers, you lose the idea of the ultra slim stick design. I think the over all SQ and cost would be much better in a more conventional system. With all the neo drivers out now, a conventional system can be made very light.

What we need are pro-sound Magna pans!

[DSM]

What we need are pro-sound Magna pans!

Here you go...

http://www.getmad.com/templates/System/details.asp?id=48555&PID=733805

[AMC]

Here you go...

http://www.getmad.com/templates/System/details.asp?id=48555&PID=733805

Those are COOL! I want some.

[DSM]

Those are COOL! I want some.

Yep.

Re Line Arrays...

1) Comb interference exist no matter what. The constructive and destructive interference is what makes the LA work. One can never get the acoustic CTCs close enough to avoid destructive lines (nulls) at the upper end of the audio spectrum. No 'pre' processing will eliminate this. A line of ribbon HF drivers... maybe but, they have little output in the low end of there BW. So now we need a 4 way box.

2) As the freq rises the line length can be shorter. There's no reason to have a HF driver in every box.

3) The polar response of LF,MF and HF must match. LF and mid band horizontal dispersion can be narrowed by using them in a doublet configuration. HF can be made wide with a diffraction slot.

4) Vertical arrays MUST be flown due to there length. The MB array must be long to work as an array.

There are solutions to all these challenges, which major manufactures claim in the sales pitch they have solved. It's just not a practical DIY system. Been there, done that...

A loud, arrayable DIY point source box would be my suggestion.

[AMC]

Yep.

Re Line Arrays...


A loud, arrayable DIY point source box would be my suggestion.

It would be the most practical. I have liked the idea of a 'W-bin' type of system with a horn lens the entire height of the box. This is not a new solution, and it seems most are doing it this way now, maybe there is a reason. Nice thing about going this way is you can start very small and just keep adding to the system for larger and larger venues.

[DSM]

It takes many boxes to make a line array achieve the needed spl. A box that makes a good line array system cannot be used as singles.

Or,

Use a point source box with good pattern control and add boxes as needed for coverage ie, 60, 120 180 degrees. Even then, to achieve pattern control down to 300hz requires a large frontal area. A tight cluster will help.

[wg_ski]

Use a point source box with good pattern control and add boxes as needed for coverage ie, 60, 120 180 degrees. Even then, to achieve pattern control down to 300hz requires a large frontal area. A tight cluster will help.

What, this? I thought that's what the OP was trying to avoid.....

[wg_ski]

It would be the most practical. I have liked the idea of a 'W-bin' type of system with a horn lens the entire height of the box.


Careful... That sounds like a DR-horn.

[AMC]

Careful... That sounds like a DR-horn.

Yep, and used by many other designs also. I have seen plans for such items from back in the 70's, and I bet they go back further then that. It's a good design, there is a reason lots of people use it.

[DSM]

What, this? I thought that's what the OP was trying to avoid.....

Nice DIY.

[Randy L]

It seems the column speaker is making a resurgence, several big name companies are now marketing new install products based on the idea including Tannoy Qflex and the JBL CBT series. This JBL design puts the tweeters in a row in front of the mid column..
http://www.jblpro.com/catalog/general/ProductFamily.aspx?FId=89&MId=2

Wg-ski. You're correct about the type of speaker I'm envisioning. I'm leaning towards a design similar to Paul's link to the new JBL speakers (see above). But rather than installed sound applications, I'd like it to be able to be used as a FOH for a small club/stage and DJ use (or one-man band).

[AMC]

This is where mass market mass production takes a seat behind no one. Even with all the collective knowledge and skills on this board, you just can't build anything like the JBL line arrays or even the Bose PAS and come out on the good end.

[Randy L]

AMC, I fear you may be correct.

But if you could design a similar design to the Bose or JBL that is 1/3 the cost, but admittedly much less pleasing to the eye, and sounds decent, would you build it and use it?

[Randy L]

Got a question.
First off, to minimize confusion of the line array term, I think we should use JBL's term of "column speaker" for this project.

I think all agree that a separate woofer is a necessity for this column speaker. If a pre-built pro passive crossover was used, rather than assembling your own, would a 500hz crossover from woofer to the column speaker be OK? I think somewhere around 150-250hz might be better, but I don't see a pre-built one offering this that can handle a good 400-600 watts RMS.

[Paul O]

I don't know about that, there's no reason a DIY box should look worse than a commercial box, The Kuze array in the Project showcase looks every bit as good as a PAS IMO and more than likely sounds as good or better. And at an estimated $1200 or so for the pair they're at least half the price of the Bose system which starts at $2600. Beating anything Bose makes in terms of performance and dollar value isn't at all hard, but doing the same for a JBL product may be another story. Still I think you can get reasonably close as long as you can get appropriate drivers and have the necessary fabrication skills.

[Randy L]

I think one of my biggest problems or handicap is making the speaker grille look nice. I would definately go for a full screen grille but not sure how to secure it and make it all look clean.

[Randy L]

OK, moving forward. Tell me what you think of these choices.

Since this will be a 3-way design (woofer is in a separate cabinet) I guess we can say the drivers in the column speaker will be mids & tweeters.

For the mids, I see a choice of either a 5" or 6" PA driver.
Dayton 5" driver, for the relative inexpensive cost keeping the overall project within a better budget.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-010

------- or -----------
Dayton 6" driver
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-015
The Eminence Alpha-6
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=290-400
The Selenium 6" woofer

For quantity, I'm thinking 8-10 drivers per column cabinet

Now the tweeters would be mounted in FRONT of the mids, like the JBL speakers linked above.

While I like the Dayton Planar tweeter, being 3-1/2" wide, it might cover the mids a bit too much preventing even dispersion from the mids.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=275-085

I found the Tang Band Neodymium Dome tweeter which looks like a good candidate for this project.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-804%20%20%20%20%20%20%20%20&FTR=264-804&CFID=25451169&CFTOKEN=50983616
Being truncated, they would fit very tightly together and wouldn't cover the mids nearly as much as the planars. Their response curve isn't as flat as the planars but it is very smooth.

Since the planars are 6-5/8" tall, I'd be using 6-7 per column.
If I used the Tang Band tweeters, I'm thinking of using 10-12 per.

Opinions or other suggestions?

I drew up a rough design on MS Word showing what the two tweeters would look like in front of 6" mids. I hope it attached to this post OK. No luck...it's larger than the 19.5kb size limit.

[Paul O]

I like the little TB tweeters, a custom aluminum or steel mounting bracket that spans the height of the cabinet in front of the mids would be needed.

If you look at some commercial cabinets they don't use any high tech method to fasten the steel grill, recently the trend seems to be to use a wrap around grill that fastens with screws in a routed recess on the sides of the cabinet.

http://www.americanmusical.com/ProductImages/Large/p42208.jpg

PE sells this perforated steel now in precut sheets.. http://www.parts-express.com/pe/showdetl.cfm?Partnumber=262-878

[AMC]

AMC, I fear you may be correct.

But if you could design a similar design to the Bose or JBL that is 1/3 the cost, but admittedly much less pleasing to the eye, and sounds decent, would you build it and use it?

I would, But I'll take hand built ugly over store bought pretty any day. But it's not just the looks, you soul able able to make a nice enough cab, but if your time is worth anything to you, very quickly the costs become excessive also. If this is a hobby, and you do it for fun, that's different.



I think all agree that a separate woofer is a necessity for this column speaker. If a pre-built pro passive crossover was used, rather than assembling your own, would a 500hz crossover from woofer to the column speaker be OK? I think somewhere around 150-250hz might be better, but I don't see a pre-built one offering this that can handle a good 400-600 watts RMS.

WAY TO HIGH. You need to go active and get it down below 150Hz, I like it at 100 or below. One problem (the one that bugs me the most) is always that lower range vocals end up in the bass box, and that sounds awful. It would also mean you have to have them with the towers, and that affects the look and sound. Build a couple of Tuba 30 or alike to go with them and you will be happy. or SCOOP, SCOOP, SCOOP (that's just food for the scoop haters) ;)

[wg_ski]

WAY TO HIGH. You need to go active and get it down below 150Hz, I like it at 100 or below.

So does everyone else who cares about the sound - but how the heck are you going to get any meaningful output at 100 Hz out of a narrow little tower? What's loud in your living room turns into nothing in a high school gym. The suggested Dayton drivers (or even the Alpha 6) don't have any throw to speak of. Use drivers with a longer throw and there goes the sensitivity. So it's either cross over too high or live with relatively low maximum output.

Whatever you end up doing it's going to be a compromise. If it were my project, I'd build it with 8's and call it good. But that's not in the spirit of things, is it?

[wg_ski]

And if you've really got something against 1" compression drivers, there's this idea. I actually considered building it once, back when a suitable 4" Tangband was on sale. You could make the three arrays any size - I would make the 4" as long as the stack of 8's, but go only as long as needed with the PT2's. 3 way offers the freedom to put the HF crossover up high enough where you won't be constantly blowing tweeters. The design is a little bigger than those little personal systems, but would still fit in Honda Accord.

[Paul O]

I don't think Randy L is suggesting this system be capable of filling a highschool gym with high spl sound.. at least I hope he isn't. It should be capable of some pretty good output.. about on par with your average 15+1 box but with better mids clarity. And I agree with AMC that 500hz crossover from the bass drivers is way too high, it has to be down around 100-150hz or SQ goes all to hell. 8 of those Dayton 5" or 6" PA drivers per side will handle that no problem, the big issue is will the stack of tweeters handle the power necessary to keep up, and can they be crossed low enough to keep the response relatively wide at the upper crossover frequency. To do that it has to be around 2.5khz.. not 5khz. As for the bottom end just about anything would do but boring old reflex boxes using 15's or 18's would mesh very well and allow some experimentation with the lower crossover frequency. And I'd want a DCX for crossover and processing duty no doubt about it, at some point down the road if the boxes turn out to not need a lot of correction(doubtful) you could build a passive crossover for them, but I'd always have an active crossover between these and the subs.

[Randy L]

Thanks for all the inputs and suggestions guys! :D

Maybe going active will be the better (and easier) way to go. I then wouldn't have to worry about designing and building a passive crossover since the pre-built pro PA crossovers don't offer the x-over points needed.

I already have an active crossover unit so no added expense there. Also, the difference in audio quality between active and passive is night & day.

Now, anyone have suggestions on how to build a two piece cabinet that locks together tightly? I don't want to have a 6-7 foot tall column to tote around. I'm thinking the bottom third would have just the "mids" and that is where the column would be split into two pieces. The top column would have the rest of the mids and all the tweeters.

[AMC]

So does everyone else who cares about the sound - but how the heck are you going to get any meaningful output at 100 Hz out of a narrow little tower? What's loud in your living room turns into nothing in a high school gym. The suggested Dayton drivers (or even the Alpha 6) don't have any throw to speak of. Use drivers with a longer throw and there goes the sensitivity. So it's either cross over too high or live with relatively low maximum output.

Whatever you end up doing it's going to be a compromise. If it were my project, I'd build it with 8's and call it good. But that's not in the spirit of things, is it?

This project is all about compromise. It will NOT be perfect, that is the only guarantee we have.

I also think 8" is where you would have to start for true full range coverage, but 6" would work as long as you understand that you will most likely have a dip from about 150ish to where ever the bass comes in, maybe at 100. And while this is not ideal, in real world applications it does not sound horrible. The old Bose AM5 systems had an outrageous hole from about 200 to 100 hz, and yet for non-critical listening they were fine, and Bose sold MILLIONS of them, made a bunch of money myself selling them, and the general public liked them. And that is really the target of the sound here, the general public. Sometimes it's easy to get distracted by OUR desires for perfection, and lose track of the fact that something like this is for a club/bar. Lets be honest, as long as there is plenty of bass, and enough mid and high to hear the vocals, it will be fine.

[AMC]

Thanks for all the inputs and suggestions guys! :D

Now, anyone have suggestions on how to build a two piece cabinet that locks together tightly? I don't want to have a 6-7 foot tall column to tote around. I'm thinking the bottom third would have just the "mids" and that is where the column would be split into two pieces. The top column would have the rest of the mids and all the tweeters.

The easy way would be to use a pole/base system that most pro sound cabs already use to raise their speakers. Just use a very small pole that would end up inside both bases. I think of bigger concern is how are you going to make them stable? I might suggest a larger base, 24 X 24" steel, that has an integrated pole in the center, going in to the base in the bottom of the bottom cab. It would be very low, use like 3/16 steel, and becomes 'invisable'. Flying them is the best way, but that make portability a problem, but putting some eyes in to allow that would be smart.

[Paul O]

Now, anyone have suggestions on how to build a two piece cabinet that locks together tightly? I don't want to have a 6-7 foot tall column to tote around. I'm thinking the bottom third would have just the "mids" and that is where the column would be split into two pieces. The top column would have the rest of the mids and all the tweeters.

I don't think the lower cab is necessary at all, simply build a pair of 4 foot high cabs with mids and tweets and put them on a stands.

[Randy L]

Agreed that a wide "foot" will be essential for stability. Already had that planned, except for material, but I like your suggestion on making it metal to lower the center of gravity.

On the idea of flying, just had a thought. I use two tripod light stands. They would hang very nicely in front of the tripod pole. I would have to extend the legs wider to compensate for the "forward bias" of the weight to keep it more stable. That might look kinda nice. Column hides the light stand and makes for a less cluttered stage appearance. But, for a (small) live band use or for single musicians this may not be feasible since their light stands are to the far sides.

[AMC]

Agreed that a wide "foot" will be essential for stability. Already had that planned, except for material, but I like your suggestion on making it metal to lower the center of gravity.

On the idea of flying, just had a thought. I use two tripod light stands. They would hang very nicely in front of the tripod pole. I would have to extend the legs wider to compensate for the "forward bias" of the weight to keep it more stable. That might look kinda nice. Column hides the light stand and makes for a less cluttered stage appearance. But, for a (small) live band use or for single musicians this may not be feasible since their light stands are to the far sides.

I have seen people use large portable systems for flying very large (maybe 12 feet long X 30 inches wide) true line array systems, so I don't think your idea is out of line at all.

(I do believe I Punned)

[Clee]

I have been researching this idea for a while, I came across this that may interest you. It is a Bose L1 knock off.

http://www.youtube.com/watch?v=R32RkxnI7oY

There are 6 video's.

Not sure about the tweeter-box at the top of his array.

[Randy L]

Thanks Clee!

[Clee]

Your welome.

Let me know what you all think of it.
I wonder about the 2 tweeters on a T at the top of it.

[Paul O]

I have seen those videos before.. some pretty amusing stuff there. ;)

[AMC]

I have been researching this idea for a while, I came across this that may interest you. It is a Bose L1 knock off.

http://www.youtube.com/watch?v=R32RkxnI7oY

There are 6 video's.

Not sure about the tweeter-box at the top of his array.

I love the 'make it cheap as possible' attitude of this guy. Interested to know how they actually sound.

Side note, Am I the only one who finds it funny when people put you-tube videos up in an attempt to show off the sound of the speakers? Between your camcorder mic, and my computer speakers, Something gets lost in the stream.

[Taterworks]

And if you've really got something against 1" compression drivers, there's this idea. I actually considered building it once, back when a suitable 4" Tangband was on sale. You could make the three arrays any size - I would make the 4" as long as the stack of 8's, but go only as long as needed with the PT2's. 3 way offers the freedom to put the HF crossover up high enough where you won't be constantly blowing tweeters. The design is a little bigger than those little personal systems, but would still fit in Honda Accord.

I really like the look of this solution. Split it in the middle and work out a way (maybe with road cabinet latches or something like that) to fasten the two halves together when you get to the gig. Put the 4" drivers in a separate sealed subenclosure. Make sure that the length of the arrays in each passband supports line array behavior down to the next crossover frequency. Then down below, use one or two dual-15 or dual-18 subwoofers based on LF need.

The only reason I think smaller ground-stacked line arrays haven't taken off is because of the stability of stacked boxes, so make sure to consider stability of the setup as well. The best-looking application of the SLS line array boxes I've seen actually mounted them to stands made of lighting truss that gave them a wide, stable base. Maybe use a length of steel pipe that runs the length of the back of the box could provide rigidity.

The other way to get away from the stability issue is to fly the things, but that's just all kinds of bad when people try to work a flown system out on their own instead of entrusting it to qualified professionals. Even a flown 12" two-way speaker box will kill a person standing underneath it if it falls on them - it does have that much mass.

[Randy L]

Rory,
Are you talking about butterfly latches like this one?
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=262-200

That would work to secure the cabinets together. I'd use them along with a "insert pole" between the two columns.

Looks like the design is starting to come together! All good things take time though.

[Taterworks]

Yes, I was talking about the Butterfly latches to secure the cabinet halves. They should have the clamping force to get the job done. Use four (two on each side) on the sides and one in the middle on the back.

For subs, if you're not planning to build your own, the Community S-Series dual-18" and dual-15" subs have their pole sockets placed closer to the rear of the cabinet, in the best possible place for reinforcing a ground-stacked line array like this.

The Renkus-Heinz IC Live system is similar to what we're talking about, except it's not a true 'line array', just a column speaker with steerable output. It goes together in about the same way, though.

[Randy L]

Fortunately I already have the "sub" so that will be an expense not needed. But...for those that are interested in building the "column" speaker like I plan to, we might want to address the sub design. I prefer using "woofer" since the crossover will most likely be between 150-200hz, which is above a typical sub's range & design.

I realized, in an effort to shed on weight, and make the whole system portable, I'll have to buy a good pro amp. It'll have to be one of the new digital switching amps and with lots of power.

[Randy L]

I just noticed something about the Tang band tweeter.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-804

Mfg specs, show a rated power input of 8 watts, 80 peak.
Having 10 of these in the speaker, crossed over at around 2500hz, how much power (RMS) do you think they can comfortably handle? I'm projecting an overall RMS capability of 200 watts just for the tweeters (and 500 watts for the mids), which would equate to 20 watts per tweeter. The low 8 watt rating scares me a bit and might change everything.

[wg_ski]

How much do you really expect a tweeter that weighs a couple ounces to handle? 8 watts thermal sounds about right to me. 10 of them might handle more load than an ASD1001.

I'd be using a light bulb on the highs for sure.

[Randy L]

How much do you really expect a tweeter that weighs a couple ounces to handle? 8 watts thermal sounds about right to me. 10 of them might handle more load than an ASD1001.

I'd be using a light bulb on the highs for sure.

That's what confused me. I know Tang Band is a very reputable company but the peak wattage being 10X the rms rating (8/80) seems "out there." I checked other TB soft dome tweeters and they all have the same 8/80 rating.

Agreed that an exciter lamp will be used to protect the tweets.

[wg_ski]

When you see peak ratings like that it means *really* short term - 10 milliseconds or so. I've seen specs on the Morels and Vifas and those can be 200-1000W - even though thermally they're good for maybe 20. The usual 2x RMS can be handled for several seconds to several hours, depending on how much thermal mass there is to heat up. High frequencies typical of what you feed a tweeter have high peak to average ratios due to the complex waveforms.

If you rated a 2" compression driver (with 4" VC) in the same way, it would be good for well over 1000 watts. And you can use them with kilowatt-plus amps on a headroom-only basis, as long as the average over a 1 second or so period is less than 50 watts. You really need a good average-responding limiter (like a light bulb) to get away with this.

[Randy L]

Here is my first rough draft of the cabinet design. Using 5" mids, the cabinet should be no more than 6-3/4" wide (outside dimension).

http://i6.photobucket.com/albums/y210/97LT1/LineArrayDrawing5.jpg

I'll just have to work out the details on how to "suspend" the tweeters in front of the mids, but I think I've got a good idea on that.

[AMC]

Here is my first rough draft of the cabinet design. Using 5" mids, the cabinet should be no more than 6-3/4" wide (outside dimension).


I'll just have to work out the details on how to "suspend" the tweeters in front of the mids, but I think I've got a good idea on that.

OMG, tweeters out 2 inches in front of your mids? The time alignment will be unlistenable, how could anyone stand to be in the same room with THAT?
:)
Just kidding, Looks good, when will you start?

[Randy L]

OMG, tweeters out 2 inches in front of your mids? The time alignment will be unlistenable, how could anyone stand to be in the same room with THAT?

Ummm did you not look closely enough? It's actually 1-7/8" in front! :p

I'd like to begin finalizing the plans--to include internal bracing, cabinet materials (wood if not a stiff polycarbonate that's easy to get at Home Depot/Lowe's), base, etc. I also need to figure the series/parallel wiring of the drivers to see how many I'll actually need to get me to my 8-ohm impedance goal. I'm thinking the total number will be at least 9 or 12 drivers each (mids & tweets).

[AMC]

Ummm did you not look closely enough? It's actually 1-7/8" in front! :p

I'd like to begin finalizing the plans--to include internal bracing, cabinet materials (wood if not a stiff polycarbonate that's easy to get at Home Depot/Lowe's), base, etc. I also need to figure the series/parallel wiring of the drivers to see how many I'll actually need to get me to my 8-ohm impedance goal. I'm thinking the total number will be at least 9 or 12 drivers each (mids & tweets).

If you are starting with 8ohm drivers, any number you can get a square root from will work, 4,9,16... they will be exactly 8ohm, Others will get you close.

[Paul O]

9 drivers? Must have been a typo there and you meant 8. To get back to 8ohms you're looking at 4, 8, or 16 drivers.

[wg_ski]

No - 9. Three in series, three groups in parallel.

[Clee]

What is the depth measurement on your proposed cabinet?

[AMC]

9 drivers? Must have been a typo there and you meant 8. To get back to 8ohms you're looking at 4, 8, or 16 drivers.

No.

[Randy L]

What is the depth measurement on your proposed cabinet?

I'll be making it as shallow as I can while maintaining the required air space per driver, which really won't be much.

I've pretty much decided to use the Dayton 5" PA driver
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-010

The specs are: Power handling: 50 watts RMS/100 watts max * Voice coil diameter: 1" * Impedance: 8 ohms * Re: 7.1 ohms * Frequency range: 90-15,000 Hz * Magnet Weight: 20 oz. * Fs: 92 Hz * SPL: 90.0 dB 1W/1m * Vas: .16 cu.ft. * Qms: 4.99 * Qes: .54 * Qts: .49 * Xmax: 2mm * Dimensions: Overall Diameter: 5-1/4", Cutout Diameter: 4-1/2", Mounting Depth: 2-1/2", Magnet Diameter: 4", Magnet Height: 1-1/16".

The tweeters have been changed due to the Tang Band's 8 watt RMS rating...I'm a bit scared about this, even if I use 10 of them.
The replacement will most likely be the Vifa soft dome. The response curve isn't quite as smooth, but pretty good.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-1024

Specifications: *Power handling: 25 watts RMS/50 watts max *VCdia: 1" *Impedance: 4 ohms *Re: 3 ohms *Frequency response: 1,320-30,000 Hz *Fs: 1,320 Hz *SPL: 94 dB 2.83V/1m *Dimensions: Overall diameter 2.75" x 2.125", Cutout diameter 1.7", Depth 1".

One thing I'm needing for my front mount tweeter design is for the tweeters to be as shallow as possible to enable mounthing them as close to the cone of the mids as possible. This will minimize and phase issues between the mids and tweets. The lowers Fs (compared to other tweeters) will enable me to lower the crossover frequency, thus avoiding the effects of comb filtering as best I can.

Since this will be for for PA use, I do not want to use any kind of L-pad or passive components that soak up and use up valuable amplifier power, to tone down the highs. All this will be achieved via a good EQ which will reduce the demand on the amp. If I can design the speaker to have a RMS capability beyond what my amp can deliver, I should be more than well protected from over driving and blowing them. The exciter lamps will still be used as protection though...just in case.

Any flaws in my train of thought?

[AMC]

Looks good. I don't think you would have any problems with the tang band, but the vifa will also do nicely, you also could try this..

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=279-234

The Fs and most other specs are about the same, and considering how many you need to buy, the cost savings could be good. And the sensitivity matches your mids.

I would disagree on the L-pad for the tweets, (with the above tweet, the sensitivity is about the same and you might not need any) but I would put it in, Make the speakers sound as correct as possible, let the EQ take care of room issues. It does make the amp less efficient in that part of the speaker, but that is kind of the point.

With 12 drivers (mids), you can make a nice 6 ohm load, that should be NO problem for an even mid range amp. Using 8 tweeters would allow you to bring them up to 8 ohms, that would also help with the level. And that give you power handling of 600W in the mids, and 200 in the tweets. that should be plenty. Just watch out, with a 2mm X-max, over-excursion will tend to take place long before 600 watts.

[Randy L]

I was wondering about the 2mm Xmaxx. But considering they will be midrange drivers, and that they will be intended for small to medium sized halls and won't be trying to compete in a SPL competition, hopefully the sheer number of drivers used will prevent them from ever hitting their limits.

Realistically, I don't see each column speaker ever seeing more than 450 watts peak, with about 350 watts to the mids and 100 watts to the tweets.

[Taterworks]

Hello Randy,

A few observations.

First of all, it looks like the design effort is going pretty well. Soon you'll need to think about crossover design. You'll want to make sure you place the crossover frequency between the mid column and HF column no lower than 2600 Hz based on the capabilities of the tweeter (you generally want to XO a tweeter at least an octave above its Fs), and also make sure you're using enough tweeters that they'll survive while you're trying to get the level you need. In fact, it looks like you were planning to use twelve tweeters, but I recommend bumping the number up to 16. Lengthening the tweeter line will also help maintain more consistent vertical dispersion (better line-array behavior) across the entire HF passband and better decay characteristics in the far field. When you voice your speakers, don't set them up in your living room. Instead, take them outside and stand some distance away -- because high frequencies tend to decay with distance more rapidly than low and mid frequencies. A large number of HF sources is a chief benefit of most line array systems, because of reduced distortion since each individual driver doesn't need to work nearly as hard.

When you design the tweeter array, it will be tempting to orient the tweeters with the truncated edges vertically to try and minimize the width of the tweeter array in front of the midbasses, but instead I think you ought to focus on mounting the tweeters with the truncated edges horizontal, to minimize center-to-center spacing of the drivers. There are two competing interests here - minimizing the HF array width in front of the midbasses, and preserving as small center-to-center distance between tweeters as possible. What I'd do is mount the tweeters to the tweeter array baffle and then use a flush trim bit with a bearing to trim along the edges of the tweeter frames so that there's no wood overhanging the edges of the tweeter flanges. The shape of the tweeter baffle will look something like this when you're done:

(__)
(__)
(__)
(__)
(__)
(and so on)

There's no need to design a trapezoidal cabinet for your arrays - they won't be 'arrayable' horizontally in the way a traditional point-source box with a horn-loaded HF is. Because of the narrow shape of the array (and the necessary slim width of the tweeter array so as not to excessively shadow the midbasses and cause internal reflections between the mid cones and the back of the tweeter arrays), these arrays will be physically incapable of providing horizontal pattern control - instead, you'll get broad coverage, maybe out to 150 degrees or so, from a single cabinet and a single coherent line of drivers. A rectangular cabinet will be the easiest to build with no apparent drawbacks (since arrayability of multiple cabinets of this design is out of the question).

If your side panels extend out past the baffle as shown in your line drawing, you'll want to use some egg-crate foam (which PE sells) bonded to the inside surfaces of the cabinet sides along the length of the tweeter arrays - otherwise, diffraction spray from the edges of the tweeter array will be reflected off the vertical surface of the side panels extending past the baffle and cause response anomalies and smear.

Check out this SLS product, which is similar to what you're doing, but somewhat smaller: http://www.nczon.com/LS6593V2.htm

[wg_ski]

With subs, you may have to handle a sustained bass boooooom. Midbass drivers don't usually see all the power the amp has at 100, 200 Hz or wherever it's crossed over at. There is usually significant content at other frequencies at the same time.

[Randy L]

Thanks to all who have contributed to this project. Your opinions and recommendations are definately helping me to properly design this speaker.

It's far from being done so be warned for more feedback requests!!

[Randy L]

Update.
Almost ready to pull the trigger on the drivers.

Need one last bit of opinion on the "mids". Which one of the two choices below do you think will produce the best, cleanest, most articulate sound?

Yea yea, it's going to be used for DJ'ing, but I also want it to have a potential for small-medium sized band use.

The Dayton pro 5"
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-010

or the Peerless SDS 5-1/4"?
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=299-247

Price is comparable, so that's a moot point.
I see the Peerless has 3mm of Xmaxx to the Dayton's 2mm, but for a midrange driver "coming in" at around 150hz, would that really make any difference?

Another thing I'm wondering about is that the Peerless has a single-roll outer suspemsion VS the Dayton's "ribbed" style. I'm thinking the Dayton driver will be more stiff lending to a sharper lower mid note. Maybe more control of the cone when being pushed hard? Keep in mind that each driver will realistically only see 45-60 watts PEAK (limit of the amp I plan on using) so I'll be well within the ratings of either choice.

[jeff_free69]

What did you decide for the tweeter & XO point?

The Dayton is a "fullrange" & its spec looks like is actually usable up to 4k (its "rated" to 15k LOL)

The peerless is a "woofer" and is rated to 6k , but theres no graph showing what it does above 1k

if you're using hifi tweeters I'd avoid crossing over at their usual 1.5k , just because I'm nervous about such things. I'd lean towards the dayton and a higher XO
but I haven't heard these drivers ...

- why not buy a couple of each and listen before pulling an triggers?

[Randy L]

Jeff,

For the tweeters, I am going with AMC's recommendations of this...
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=279-234

It's on sale right now, and at $5.46 each, it's almost too good to pass up! Each column will have at least 10 of them.
Its Fs is 1400hz, but I'm not sure why its rated response says 3000hz on the low end. The response graph shows a flat response starting at 1500hz.
The crossover point will most likely be around 2khz (no higher than 2.5khz), 2nd order.

The Dayton 5" has a lot of great reviews and I do trust the Dayton brand in delivering a quality product. Looking at its response graph, the 5" appears to break up above 3khz. I wouldn't consider this a full range driver, though.

I haven't been able to find any response graphs for the Peerless driver, but another site states this model number has a super flat response and was made for the Miller & Kreisel speaker mfg, widely known for very high end models. My reasoning tells me, then, that it must be better than the average driver.

[jeff_free69]

OK looking more closely at the dayton again - its probably fine out to 3 - 3.5k. the break up is more like 5.5k +

I agree that I'd expect the Peerless to sound better just because of what it is. That doesn't mean it doesn't have any peaks that might need notching, though.

[Randy L]

For some reason, my gut instinct is telling me to use the Peerless driver. :confused:

[Texan]

My vote would be for the Dayton
1 It has a pro type surround that might be a little tougher
2 It has lower vas for a smaller cabinet volume
3 It has lower Q for smaller cabinet volume
Hopefully others will confirm or deny these as good selling pionts

[Randy L]

Thanks for your input Tex. I wonder. though, if the Q of the Dayton being so low (.46) that it's better suited for vented enclosures, whereas the Peerless' Q of .63 makes it ideal for closed enclosures (which this design will be).

On the other hand, the Dayton will enable a much smaller enclosure size, like you said, than the Peerless. But since I won't be using this for bass reproduction (noting the crossover of around 150hz to my 15" woofer in separate cab), would the Peerless be able to sound great in a smaller enclosure (of say .2 s.f.)?

[Texan]

Normally I would agree, but I look at it this way. A high Q means no good for vented, but the opposite is not necessarily true. Low Q is ok for sealed it just means you won’t get as much bass extension as with vented. I think this is ok for your app since in 0.07ft^3 per woofer f3 is below 150hz. The peerless will get your f3 below 100 so you might be able to cross lower, but the pole piece is not vented so that might be another thing that doesn’t lend itself to longer durations at higher levels. The peerless will require about 5 times the volume.:eek:

[Randy L]

The peerless will require about 5 times the volume.:eek:

Tex, can you elaborate on your statement? I see both drivers have the same sensitivity of 90db SPL @ 1W/1M.

[Texan]

Tex, can you elaborate on your statement? I see both drivers have the same sensitivity of 90db SPL @ 1W/1M.

Oops sorry
Volume as in cubic feet or liter etc. inside the enclosure

[Randy L]

LOL...sorry. I should have noted that you were talking about airspace needed. My bad. :p

I ran some box size calculations and the peerless--in my column design--would require the depth to be pretty darn deep. This is really the only variable I have since the design criteria is that the column speaker be as thin as possible. Also, the line array "properties" will necessitate the drivers to be placed as close to each other (vertically) as possible. So the "required" width will be about 6" max, and the height would be 5-1/4".

With that said, it apprears the Dayton will be the best choice. Thanks for the "tip"!!!!

[Randy L]

http://www.parts-express.com/pe/pshowdetl.cfm?&DID=7&Partnumber=279-234&ctab=3#Tabs

Before I pull the trigger and place the order, I need help on determining the crossover point for this tweeter. The specs listed show a freq response of 3khz, but Fs is 1400hz, and the mfg's response curve shows the tweeter is flat down to 2khz. I think Peerless is using the golden rule of crossing over at one octave above Fs (2800hz).

Only problem is that I need to cross it over no higher than 2600hz to minimize the effects of comb filtering, right?

With that said, do you think a 2500hz, 2nd order, crossover point (with exciter lamp protection, will be enough to protect the tweeters? Keep in mind that that there will be 10 or 12 of them and I highly doubt that they will see more than 25 watts peak each. (my best guess on RMS for music program will be 15 watts, considering the highly compressed nature of recorded music these days).

[Randy L]

Amazing how things suddenly pop up to throw a wrench into plans.
I don't know how I missed this tweeter before, but I have found another that I think will be perfect for my goals and design parameters.

It's the Vifa 1" Soft Dome.
http://www.parts-express.com/pe/pshowdetl.cfm?&Partnumber=279-168

Doing paper calculations, 10 of the original tweeters I was going to use (the Peerless) would have got me a little over 96db (1W/1M) and 250 watts RMS.

The Vifa, with its 94db SPL (1W/1M), 10 tweeters will get me just over 103db SPL. Almost 10db gain for the same amplifier power! It's 50 watt RMS rating is WAY more than I'll ever feed it....allowing me a 500 watt RMS capability. Wow.

Bonus is that the Vifa's rating of 1350hz on the low end (also its Fs), will be almost perfect for my intended 2.5 or 2.6khz passive crossover point.

++ bonus is that they are almost identical in mounting design so I will not have to modify much--if anything.


So they'll be mated to 12 of the Dayton 5" pro drivers
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-010

My calculations show a 1W/1M SPL total for 12 of the Daytons to be roughly 100.5db SPL.

[Mayhem13]

Either tweeter will be fine with your intended XO point Randy as the low end isn't too much of a concern since there's multiple units....that's just one of the plusses of line arrays. I'd go with the Vifa as in the unlikely need for a replacement or multiples, you'd be able to do so. I'd look for a midwoofer in current production as well instead of a buyout for the same reason. I haven;t read all of this thread but i'm assuming commercial duty so plan ahead.

[Randy L]

^^ Yes on the pro application. Targetted for DJ's and small bands in small to medium sized venues.

[Randy L]

OK...I need everyone's help to make sure I'm looking at the wiring of the drivers correctly, as well as the power each speaker will be seeing.
I made this wiring diagram up which shows the individual driver impedance as well as the (anticipated) power provided to each row of speakers.

http://i6.photobucket.com/albums/y210/97LT1/ColumnSpeakerWiringDiagram.jpg

LEGEND:
-- Each row indicates those speakers are wired in SERIES
-- Each row joined by a line are rows wired in PARALLEL
-- the number in each square (speaker) is the impedance of each
-- the number above/below the box with a "W" after it is the wattage that speaker will see when the specified total wattage is provided to that row.

[Randy L]

Anyone see a problem with this configuration? If not, I'll just need to get the proper passive crossover components to build my own and then order everything.

[rickcraig]

We built a CBT array for our church (seats about 1,000) and it works great. If you can attend the Dayton festival in July you can find out more.

[Kevinized]

Anyone see a problem with this configuration? If not, I'll just need to get the proper passive crossover components to build my own and then order everything.

I prefer a system with no HF crossover. I believe HF crossovers introduce phase distortions.

I used these drivers in the system I use as a stage monitor.

http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-378

I'm in the process of building Version 2 with 4 drivers (to lower coil resistance), and 2 cabinets with 8 drivers each to use as mains.

Each speaker will be self powered so set up will require placing the monitor, plugging in the instruments and power and I'm ready to go.

[billfitzmaurice]

I prefer a system with no HF crossover. I believe HF crossovers introduce phase distortions.That's the main argument made for single driver systems, and it has a ring of truth to it. But there's never a free lunch. Where single driver systems are concerned you either sacrifice low end by using a small enough driver to go all the way up with adequate dispersion, or you sacrifice HF dispersion with a driver capable of going all the way into the low end. In a stage monitor where neither deep lows nor widely dispersed highs are necessary you can get away with a single driver. For mains two ways are very much the minimum required for a good full bandwidth result.

[Music is life]

Randy that wire diagram looks correct.

[Randy L]

OK, so looks like 9 of the Daytons it will be. At 5-1/4" outside diameter, nine aligned on top of each other gives me a vertical dimension of 47-1/4". This will fit perfectly into my plan since I can easily carry and load something 4' long.

Since the objective is to have the top driver at 7 feet high (about the same height as the Bose L1), I'll have to make a base for my columns. That's easy enough, PLUS, it benefits me in allowing all drivers to be in one cabinet rather than splitting them up. It'll be the same design as the "top" to make it look like one piece and latched together using those twist-lock PA cabinet latches.

I will not worry too much about the width nor depth since I'll be fine tuning the design to fit the drivers. But they will be as small as I can get them. So I'm forecasting an outside width of 7-8" and an overall depth of 8".

Now...need opinions on this. Should I pack the inside with acoustic padding or leave them empty? Since I'm using void-free Birch plywood, I don't think I'll have the same resonance problems as other similar projects using PVC.

[billfitzmaurice]

Should I pack the inside with acoustic padding or leave them empty? Since I'm using void-free Birch plywood, I don't think I'll have the same resonance problems as other similar projects using PVC.:confused:
The need for damping is only peripherally related to the cabinet material. The only cab alignment where it's not a required element is a folded horn subwoofer.

[Randy L]

I'm thinking I won't need any since the stiffness of the Birch (although it will be 5/8" not 3/4") and the lack of a relatively large "plane" of unbraced wood surface area will keep resonances down. Not 100% sure since this is my first foray into ths type of design. We'll see I guess.

[billfitzmaurice]

I'm thinking I won't need any since the stiffness of the Birch (although it will be 5/8" not 3/4") and the lack of a relatively large "plane" of unbraced wood surface area will keep resonances down. Not 100% sure since this is my first foray into ths type of design. We'll see I guess.
Damping is not used to control panel resonances. It's used to absorb internal midrange reflections and, in the case of a low frequency enclosure that's at least 1/4 wavelength in dimension within the passband, to damp the 1/4 wavelength resonant mode and its harmonic modes.

[robwest]

Not sure if you've selected your drivers yet, but this DOTD seems like a candidate. http://www.parts-express.com/pe/showdetl.cfm?Partnumber=295-010

[Randy L]

Oh yea...that's the driver I've been waiting to go on sale. Placing the orde for a bunch today! :p

[Randy L]

Damping is not used to control panel resonances. It's used to absorb internal midrange reflections and, in the case of a low frequency enclosure that's at least 1/4 wavelength in dimension within the passband, to damp the 1/4 wavelength resonant mode and its harmonic modes.

Bill, thanks for the clarification. I just learned something new!

[Randy L]

Good things come to those who wait...pulled the trigger and ordered the drivers. Saved 40% off the Dayton 5" drivers. NICE!!!!!

Went ahead and ordered 24 of them to have spares handy. If this design exceeds my expectations I will use them as my mains for my home theater system when not being used for my gigs. Then, I can use the extra drivers to make a voice-matched center channel using 4 Daytons and 4 of the Vifa 1" tweeters.

Sawdust will be created soon...but the Vifa's are on backorder.

Haven't yet decided on the 2-way, 2nd order, crossover components. I'll need to call tech support for the optimum parts for this high power system.

I'll also be ordering a new pro amp...a Carvin DCML3800 to drive these.
http://www.carvinguitars.com/products/single.php?product=DCM3800L
A ton of power in just 2 rack spaces and 15lbs, for just $800. This one amp will replace two of my existing Carvin DCM2000's (each weighing 34lbs). NICE!

[wg_ski]

I'll also be ordering a new pro amp...a Carvin DCML3800 to drive these.
http://www.carvinguitars.com/products/single.php?product=DCM3800L
A ton of power in just 2 rack spaces and 15lbs, for just $800. This one amp will replace two of my existing Carvin DCM2000's (each weighing 34lbs). NICE!

I sure hope it's more than just a pure class AB design. SMPS or no, that would generate an awful lot of heat. Maybe I missed it somewhere but the specs and description didn't say anything about any reduced-dissipation operating class. Remember how hot those old Macrotehcs used to run? And they has some mass to heat up. If that's 15 pounds of class G or H, it wouldn't be bad at all for $800 (compared to $1100 for the 21 pound PLX).

[Taterworks]

Randy, I'm not sure if this has already been asked, but what subwoofers are you planning to use with these?

[Randy L]

Ski, thanks for the "tip" on the amp Class. I'll research it a bit more.

I found out that the Vifa tweeter is out of production now, though there were still some left available from the mfg. Won't be available from mfg until Sept, so I sustituted with Vifa tweeter #264-1028.
http://www.parts-express.com/pe/showdetl.cfm?Partnumber=264-1028&FTR=264-1028&CFID=25451169&CFTOKEN=50983616

Specs are very close to the original Vifa. Only issue now (since the new Vifas are a little deeper) will be that I'll have to mount the tweeters out just a bit further in front of the Dayton 5"ers.

This way, if anyone likes the project and wants to build it, at least the drivers will still be in production.

Rory, my current woofers are custom built cabs using the Eminence Kappa 15-LF. They are a 3-way setup, with the two 8" mids and a compression horn in their own smaller cabinet mounted above the woofer cab with a speaker pole. Excellent bass and musicality. The passive 3-way crossover sends 500hz and below to the woofer. But this new project will utilize an active crossover to split the signal somewhere around 120-150hz. The Dayton's go down to 90hz, so I'll play around to find the best crossover point that provides the overall sound without stressing the Dayton's on their low end limits.

[edit]Forgot to add....at first, I'll use the woofer cabs as is. But, they are 46lb each and I'm going to see if I can come up with a smaller/lighter design without sacrificing too much low end. But...I'm thinking that even 18" sub cabs can be used effectively crossed over at 120hz.

[Randy L]

Found a site that has a peak SPL calculator.
http://myhometheater.homestead.com/splcalculator.html

Running some numbers through it, I see that the 9 tweeters (at the 200 watts peak I am estimating I'll put through them) will produce an SPL of 127.5db

The 9 woofers, at the 600 watts estimated, I see 127.3db. This appears to be a very well balanced system with the tweeters not overpowering the 5" Daytons.

Driven to their peak rated limits (100 watts), the peak output of the Daytons would be (9each X100 watts) 129db SPL and the tweeters would be 133db SPL.

Missed my SPL target, but still a VERY loud speaker it will be. The true test will be the listening test once done. I hope that my fellow DJ friend will be curious enough to bring his Bose L1 system so we can do a A/B comparison test.

[setzer808]

Can't wait to see the results of this. I've used my DIY line arrays in an auditorium in a typical (non-dance) PA situation. I found the results very favorable...even mostly preferable to the Bose sticks on the wall.

The youtube video was interesting, I too used fencepost for my enclosure, but didn't need the front baffle as my speakers fit in the post. I think the posts really help the appearance...that and my woodworking results still scare me.

Keep in mind I used $1.25 speakers intended for use in a TV set...but I can see small "pro-sound" style drivers really taking such a speaker to a new level!

My build: http://techtalk.parts-express.com/showthread.php?t=218607

[Randy L]

It will take a little time to get a working model since I will be doing some test fitting and mock-ups to fine tune the cabinet design. I want to end up with a design that will be relatively easy to build using a table saw, circular saw, and jigsaw, so others won't be intimidated to build it (if the final outcome proves to be successful). No dovetail joints or other complicated methods. Might end up needing to use my router though.

Plus, a little busy with work and weekends to devote uninterrupted time for this project. Hopefully I'll get the first working cabinet within the next month.

[LongTheTiger]

I know you already bought the drivers but i was actually thinking of a similar project for weddings and speaking events.

i decided that i didn't like any of the full ranges for the task but i had a pair of jbl gto-528 5 1/4" car audio coaxial laying around so i set them up on smaart to have a listen and they were great! AND the titanium hf driver in the middle pivots so you could adjust coverage in a staggered pattern live the bose cabs.
I plan on building an array (or Column) of 8 to see what i can get 90db *8 *50w so almost 125db out of 1 unit. (i could do 2 units for a 8ft stack @ 128db)

I'm looking forward to seeing your results. as i was not impressed by the pa-130 in my testing.

Good Luck!

[Randy L]

Hey Tiger,
While brainstorming driver choices, I did think about co-axials. But I got "scared away" by the possibility of considerable comb filtering by having the tweeters so far apart from each other.

As it is, I will be shaving off the tops & bottoms of the tweeters' round faceplates, which will result in a truncated shape. This should get me a center-center spacing of the tweeters at 2" to a maximum of 2-1/4".

I'll be testing all the speakers this weekend since I must make sure all tweeters are working before I modify their faceplates (no returns then!).

[captainobvious99]

Updates.....??

[ewetho]

A modification of Bills design.

http://www.youtube.com/watch?v=z4InmiT2Uzo&feature=player_embedded

or

http://billfitzmaurice.info/forum/viewtopic.php?f=13&t=13461

[Randy L]

Update....sad. As I was making considerable sawdust, things really took a turn with my job and other things which put the project WAAAAY on the back burner. I got the cutouts for the 5" Daytons, but did realize the tweeters I got would require them to be "stood out further than i anticipated originally (due to the heat dissipating fins on the tweets). One way to rectify this is to mount the mids from the back of the front baffle, but I'll still have to mount the tweets a bit further out to clear the cutouts for the mids. This will ensure all the tweets are mounted next to each other minimizing the comb-filtering.

After some playing around and test fitting, I found that a 15* angle for the sides of the cabinet seemd just about perfect. Just be VERY careful of how you set up the board when you set your table saw blade at a 15* angle for the cut.

Also, the entire cabinet, due to the many 5" Daytons, will have a consideralbe weight. A bit more than I had hoped for, but still not enough to make it impossible to move and lift up with one person.

Now that things are slowing down for me, I hope to get back to assembling everything and running a sound check. I did take some pics of the cutting and fitting of the pieces, and I can at least post those up a a couple weeks.

[Taterworks]

The weight will give it the feel of quality. I'm looking forward to pictures.

[Randy L]

Seeking opinions...originally my plan was to mount the tweeters in front of the top 4 midrange drivers. This will require the tweeters to stick a bit out, as mentioned in an earlier post. The overall height of the entire cabinet is 4' tall. Due to the complexity of devising a mounting structure for the tweeters with minimal blocking of the 3-4 midrange drivers behind them, I am considering a cabinet redign to make mounting the tweeters easier. Would it affect the dispersion any if I mounted all the tweeters in the same vertical line at the top of the cabinet, above the top midrange driver? This would add an additional 12-14 inches to the overall height.
My goal is to make the design easy to build, but not sacrifice audio quality. The tweeters will still have a very small CTC distance from each other to offset comb filtering.

[AMC]

Would it be easier to mount them on the side? It would seem that adding height to an already tall array would be more problematic then it's worth. They could even be in a separate 'add on' cab. I wouldn't be worried about blocking a few of the mid drivers slightly, the midrange that would be blocked is the strength of the driver and could easily be compensated for, if it needed it.

Also looking forward to pictures.

[Randy L]

Would it be easier to mount them on the side?

Yes it would be easier, but my design goal is to make it slender..basically to compete audibly AND visually with the Bose L1 PAS. As it is now, it will be 7" wide at the front with a 15 degree trapezoidal angle towards to back....already twice as wide as the Bose, but it should still look pretty nice, especially with the custom full faced metal grill I will be ordering.

Still scratching my temple on how to design the base for stabilization. I doubt it will be free-floor standing since it will be 4-5 feet tall and that's too low except when on a 3-4 foot foot high stage. Maybe a tophat pole and cup to be mounted on top of a woofer and a bracket enabling vertical suspension on a truss for flexibility in mounting applications.

[AMC]

Yes it would be easier, but my design goal is to make it slender..basically to compete audibly AND visually with the Bose L1 PAS. As it is now, it will be 7" wide at the front with a 15 degree trapezoidal angle towards to back....already twice as wide as the Bose, but it should still look pretty nice, especially with the custom full faced metal grill I will be ordering.

Again, the genius of Bose. They even have you designing for looks!:) I would stick with your original idea. It was not a bad idea, and I don't see any obvious benefit from changing, so... why change?


Still scratching my temple on how to design the base for stabilization. I doubt it will be free-floor standing since it will be 4-5 feet tall and that's too low except when on a 3-4 foot foot high stage. Maybe a tophat pole and cup to be mounted on top of a woofer and a bracket enabling vertical suspension on a truss for flexibility in mounting applications.

What if you built a base, that matched the face of your array, but maybe expanded a bit towards the bottom for stability. Would add a 'complete' look to it, and you could make it as tall as you think it needs to be, could even build a couple at different heights. Add some removable weight to make it extra secure, and a couple of latches to keep it from toppling over, and your good to go! Make them hollow with a door, and you have a place to put extra cabling and such.

[Randy L]

^^ Great ideas to ponder! Thanks. I gotta tell my wife that I'm through with weekend functions with family & friends for a while so I can make more sawdust... :rolleyes:

[CWH&R]

Wife mnust have put her foot down, No more reports!