Small PA line array - diyAudio
Go Back   Home > Forums > Loudspeakers > Full Range
Home Forums Rules Articles diyAudio Store Gallery Wiki Blogs Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 4th July 2016, 12:24 PM   #1
diyAudio Member
 
chris661's Avatar
 
Join Date: Feb 2008
Location: Sheffield
Default Small PA line array

So, having built some fairly powerful PA speakers*, I'm also going to build something much smaller and lighter, for those smaller gigs where you just need to reinforce acoustic guitar and vocals in a coffee shop.

Having looked around online, lots of people are using tiny line arrays with around 6 full-range drivers and a small sub. Bose L1, RCF Evox, that sort of thing.

Guess what I'm planning on building.

I happened to have 4x Bose 2.5" full-range drivers around. They're from the old SoundDock series. I found them to sound decent enough in that plastic enclosure (given by family - the iPod dock was broken) that I decided to keep hold of them, maybe for a little PC system or something. Anyway, I put all four in a cardboard box, taped them in place, wired them all in parallel to one side of an EP1500 (1ohm load, since they're 4ohm drivers), and had a listen. The result was pretty good. Not something you'd want to keep around, but there's definitely potential there. Quite a lot of power handling, except for at LF where (I suspect) air leaks produced lots of cone excursion but little below 100Hz or so. Through the mid-high range, quite a lot of output was available. To test further, I put a 12" in a ported box on the other channel of the EP1500. There was something of a sensitivity descrepancy, but once that was sorted, a ~200Hz crossover gave the little FR units more power handling.

Click the image to open in full size.


I did note that moving off axis along the length of the array produced interesting effects: when your ear is no longer in-line with one of the drivers, HF output drops hugely. This led me to investigate the minimum number of drivers that would be needed to ensure that people stood up or sat down, and have them all in the vertical coverage. Measured my head height sat down, stood up, did a bit of rounding and ended up at 12 drivers.

This was enough for me to go and find 8 more drivers.

Click the image to open in full size.
A couple of the frames were bent in transit, but the cones move freely so I'm confident they can be straightened out without breaking anything.

12 drivers is quite a long line, so, in order to contradict my findings about moving off the end of the line with the 4-driver array, I fired up The Edge, which is a rather nice piece of software, to simulate what the long array will be doing at different distances and listening positions.
Here's the 840mm array (12 drivers, 70mm square frames) at 8m distance:
Click the image to open in full size.

Pretty good, if you ignore the baffle step effects. 8m is pretty far away for these gigs, though, so lets get a little closer. 2m:
Click the image to open in full size.

Yuck. We've got 10dB difference from 1kHz to 6kHz. Well, I've heard about frequency-tapering arrays. That is, reducing the line length as frequency goes up.
Lets disconnect the top and bottom three drivers, so we're running 6 drivers
Click the image to open in full size.
This is at 8m again. Better in the HF than the long array at the same distance.

Lets try closer. 2m:
Click the image to open in full size.

Much better than the full-length array at HF. To achieve something like this, I'd add inductors to roll off the HF, starting around 1kHz). The phase shifts might do interesting things to the vertical polar pattern, though.

So, it looks like there's a balancing act to be done here. It seems that you need to be standing ~10x the line length away before everything sums nicely to give a good HF response. Tapering the array might help, but (and its a big but) the vertical dispersion is tight either way, and moving from sitting to standing would produce huge changes in the sound. I think some experimentation will be needed here. Maybe I could run only a couple of the drivers into the HF, and roll off the rest...

Anyway, that's where the experimentation will be.
Once I've got something I'm happy with (if I end up happy with it at all), I'll get myself a powered sub that has satellite outputs and an active crossover, and call it good. I might put a Behringer DEQ2496 in there if any serious EQ is needed, but I'm hoping to get away without it, just using a little notepad mixer for those tiny gigs. I'll test for SPL with the 12"s or 15"s I have to get an idea of what'd be required, but there's a few models (8", 12", 15") that I could go for. The array will be wired 3 drivers in series, two groups in parallel, giving a pair of 6ohm loads, which should be nice and easy for the built-in amps.

That's all for now. Interested to hear what you all think. The resale on these drivers is pretty good, so if I decide this isn't for me, its only time and wood used.

Chris

*Chris661's PA system thread
__________________
My work: www.grimshawaudio.com
  Reply With Quote
Old 4th July 2016, 04:33 PM   #2
diyAudio Member
 
perceval's Avatar
 
Join Date: Feb 2014
Location: Taiwan
You're experiencing the inherited characteristics of line arrays.

Vertical dispersion is pretty much limited to the height of the arrays.
The drop in HF is called comb filtering.

Here's a good read on line arrays that will make you understand why the things you measured are happening:

James Griffin's
http://www.audioroundtable.com/misc/nflawp.pdf
  Reply With Quote
Old 4th July 2016, 05:17 PM   #3
diyAudio Member
 
chris661's Avatar
 
Join Date: Feb 2008
Location: Sheffield
Hi,

I've been through that paper a couple of times. I'd like to use an array using full-range drivers. The LF is taken care of, so I don't need the extra power handling from going to a 3-way design. The arrays covered there are designed for near field use, too. Far-field will have its own set of compromises.

Having connected it all up again, the problem I'm finding is that the HF response varies considerably with distance, over the distances in which this system might be used.
Used the DEQ2496 and an ECM8000 to force a flat response, and it sounded pretty good. Moved a few metres further back, and suddenly the treble is excessive.

EQing the treble so it's more palatable at the further distance, and nearer sounds a little dull, but not too bad. Needed around 6dB of cut.

This project will be very much a "feeling my way around" sort of thing. I like the design idea - something tallish and very slim that can go surprisingly loud. There are lots of compromises to be made, and I plan to find some of them, and perhaps even work around some of those. I could, for instance, frequency-shade the whole thing, aiming for maximum power handling while ensuring the array remains effectively a point source, by rolling off drivers higher and higher until the top one or two drivers, which run full-range. Might not be enough HF power handling, but we'll see.

Chris
  Reply With Quote
Old 4th July 2016, 07:46 PM   #4
ra7 is offline ra7  United States
diyAudio Member
 
Join Date: Feb 2009
I suggest the original paper on line array by David Smith. It is fantastic.
Constant Beam Width Transducers line arrays
  Reply With Quote
Old 4th July 2016, 10:02 PM   #5
expert in tautology
diyAudio Member
 
bear's Avatar
 
Join Date: Apr 2002
Location: New York State USA
Probably, but not for certain, you're way better off letting some other driver handle the highs - like a proper compression driver + horn. Or a multi-cell, or an arrangement like that, given that you are talking "small" rooms.

Also the BOSE drivers probably do not have sufficient power handling and sensitivity to get your the SPL you would want.

Remember too, that BOSE usually EQ's the carp out of their drivers.

I'd guesstimate that you'd get reasonably good sound with a roll-off in the 2.5K range, handing over to something with proper dispersion for the highs, up top of the array, and then adding in some EQ to get it flat down to ~200-250Hz or there abouts.

More needs to be defined WRT the project, like size of the room(s), max SPL required, bandwidth required, power available, physical placement(s) available, and the WEIGHT that is transportable.
__________________
_-_-bear
http://www.bearlabs.com -- Btw, I don't actually know anything, FYI -- every once in a while I say something that makes sense... ]
  Reply With Quote
Old 6th July 2016, 04:06 PM   #6
diyAudio Member
 
chris661's Avatar
 
Join Date: Feb 2008
Location: Sheffield
From what I can tell, there's quite a lot of these little line array PA systems in use. The Bose and RCF ones I linked are a couple of examples - there's lots more.

These particular Bose drivers sound pretty decent. I wouldn't have kept them if they were anything like the drivers they use in the 901s. I've also ordered a couple of spares, and will be conducting some power handling tests. I'd guess in the region of 10-15w, and of the order of 85dB@1w sensitivity. Given I'll be using 12 of them, that's quite a gain in power handling and sensitivity, around 95dB@1w, and 100-150w power handling. Should be a match for an older-model 12" top speaker in terms of output, but in a smaller package.

I plan on buying an active sub that can drive satellites, so I will make sure power available is adequate. Weight is flexible. Less is more, but I can move 90lb cabinets alone, and I don't think these will come close to that.
The rest of it, I don't mind so much. I haven't played around with speakers like these much, so it'll be a learning experience as much as anything else.
I do have some small compression drivers and horns around, so might give them a try once I've knocked together a prototype.

Chris
  Reply With Quote
Old 11th July 2016, 06:11 PM   #7
diyAudio Member
 
Join Date: Jan 2015
Watch the excursion on those little drivers. I'd be surprised if they would handle full power even at 200Hz. Depends how much power you actually need I guess!
  Reply With Quote
Old 13th July 2016, 03:46 PM   #8
diyAudio Member
 
chris661's Avatar
 
Join Date: Feb 2008
Location: Sheffield
I'd be surprised, too.
I'm planning on putting some DSP on the front end of this setup, so I'll be making sure to set some limiters and/or dynamic EQs.
Maybe I could open the active sub and get the XO up to 250Hz...

I've been working a lot recently, but found a couple of hours to pick up a Dremel, vertical cutting guide, and lots of 90mm wide strips of 5mm MDF.
Unlike my previous projects, this one will, unfortunately, be painted black.

Chris
  Reply With Quote
Old 14th July 2016, 03:01 PM   #9
diyAudio Member
 
chris661's Avatar
 
Join Date: Feb 2008
Location: Sheffield
Tested one of the drivers to destruction today.

These are 4ohm drivers (DCR is ~3.2ohm), and the test unit was free-air.
Set off at 6v, 80Hz. Lots of cone excursion. I realised this might be somewhere around Fs, so I moved the frequency up to 250Hz, which ought to be in the low-impedance range.

I set off at 8v, which was fine. 10v, fine.
12v increased to 15v over 5 minutes or so, still fine. Sat 15v for 5 minutes or so. Still fine, though the magnet was getting warm (~30C). Warm glue smell when you get close to the driver.
Remember this is with a sine wave test signal.
I got impatient, and did a (relatively) high-power test, without letting anything cool down.

20v RMS sine wave.
52 seconds later, we had smoke. 30s later, no more noise, and a hot magnet.

Cone excursion got to around 9mm p/p, but there was quite a lot of mechanical noise. These are linear around 5mm p/p, tops.


Given its a 2.5" driver with a 0.75" voicecoil, I was very impressed with how much power this unit stood up to. I'd say its safe to take 10v sine waves pretty indefinitely, which is around 25w. Pink noise has a 3dB higher crest factor than sine waves, so this would be a 50w pink noise rated driver (albeit not tested for the 2 hours or so that the AES standard tests require).

I'm using 12 of these drivers in the column, so thermally I'd be safe occasionally clipping a 600w amplifier, though that amplifier driven with full-power sine waves would burn the lot. That's to say nothing of power compression.

The active sub I'm looking at for this setup is this one:
the box CL 112 Sub MK II - Thomann UK
I expect it'll sound alright, though not fantastic. With some EQing and perhaps a different driver (got some 500w Fane 12"s lying around), I can probably get some decent sounds out of it.

Another avenue to explore would be whether to go sealed or ported. A few holes drilled in the back (perhaps doubling up the material thickness) would put tuning in the 120Hz range, which would certainly gain some mechanical power handling, and also some ventilation.

Things to do. I'll try to make a prototype baffle soon.

Chris
  Reply With Quote
Old 19th July 2016, 02:50 PM   #10
diyAudio Member
 
chris661's Avatar
 
Join Date: Feb 2008
Location: Sheffield
Here we go...

So, after buying some small clamps, this build flew together. It ain't pretty, but it works for now. There's 2x NL4 inputs, so I can drive each series group of 3 drivers individually. I'm using a DCX2496 and an NU4-6000 for this, so I can add any crossovers, delays, EQ etc I want.

First of all, I ran all the drivers full-range. After messing with REW for a bit, I got some sensible graphs. Guess what, it looks like a 901 driver's raw response. Maybe not quite that bad, but not far off: a big hump around 1kHz, with 100Hz and 10kHz both 20dB down.
I'm going to attribute at least some of that to the line array. These drivers definitely were better than that in the Bose-designed cabinet.

Got REW to EQ the whole mess flat, and had a listen.
Where the mic was, it actually sounded quite nice. Good amount of headroom, flat, even response. Walked to the far end of the garden, and it was awful. A lot of treble. Yuck yuck yuck.

Next up, I put a low-pass filter (700Hz-ish) in on the DCX2496. This was a simple first-order filter, on the bottom 9 drivers. When I come to build the passive version, I don't want to have to mess with complex crossovers, so I kept it as simple as possible. Ran REW again (a mere 16dB between the top and bottom of the bump this time), and dialled everything into the DCX.

The result was actually really rather good. The sound stays pretty even when walking around. Perhaps a slight change in the last octave when you move over large distances, but nothing objectionable.
So, I went and got a 12" in a box, and sat back to enjoy the music for a while. Crossover LR4 at 200Hz.

Click the image to open in full size.

Click the image to open in full size.

That's all for now.
Chris
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Inexpensive Line Array for PA use? audiothings PA Systems 2 30th April 2011 12:23 PM
Line array for PA use, part II dave251 PA Systems 2 7th November 2010 02:14 AM
Small Format PA Line Array Omeone PA Systems 9 21st August 2010 05:22 PM
PA Line Array Advice Omeone Full Range 2 20th May 2010 06:05 AM
Line array for PA use dave251 Full Range 6 19th May 2010 10:21 AM


New To Site? Need Help?

All times are GMT. The time now is 04:17 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2017 DragonByte Technologies Ltd.
Copyright 1999-2017 diyAudio

Content Relevant URLs by vBSEO 3.3.2
Wiki