I understand you well, we have fairly similar philosophies on this. I too have many times rolled my own solution, either because I didn't have that much money to spend, or because I didn't think the commercial product was worth the asking price. Right now my small spare P.A. system/acoustic guitar amp consists of the electronics from a $100 Powerwerks PA-50, mated to a $5 thrift-store vintage Sears speaker. The combination is better sounding than the Powerworks was, and only cost $5 more, and an hour or two of my time.
That was all I wanted to do, to let you know that magic wasn't guaranteed. I do understand the appeal of the concept of a minimal enclosure, no crossover networks, and magical dispersion properties.
Then I got deeper into the details, and found out that you can't EQ away the bad sound of cone breakup, it takes a lot of 2.5" drivers to equal the area of one 12" woofer, and the magical dispersion properties are not so magical in practice.
Some quick numbers: re the supposed large radiating area of a line array, the piston area of one of those little 2.5" speakers is equivalent to a circle with a diameter of only 2.09" (I took Sd from the data sheet and computed the equivalent diameter). So it would take thirty (30!) of those 2.5" speakers to nearly equal the piston area of one 12" speaker.
(Here, I assumed the piston area of the 12" speaker would be equivalent to a 11.5" diameter circle.)
How about the dispersion advantage? Well, the treble dispersion from a speaker starts to suffer if the diameter of the speaker is bigger than half a wavelength. In the case of a piston 2.09" in diameter (that's 5.30 cm), this happens at a frequency of about 6.45 kHz. So we can expect our 2.5" Tymphany to start to beam treble above about 6.5 kHz.
Taking a quick look at the data sheet, sure enough, the 7 kHz response is about 7 dB down 60 degrees off-axis.
Okay, let's look at frequency extension on the bass end of things. The little Tymphanys have a resonant frequency around 140 Hz - enough for female vocals, ukuleles, mandolins, fiddles. Maybe enough for male vocals. Guitars would lose most of an octave on the bottom end. Forget keyboards, cellos, and other instruments with any substantial amount of bass energy below 140 Hz.
The laws of acostics dictate that life is hard for a 2.5" speaker. It's too small to do a good job with bass, and too big to do a good job with treble!
I have a whole and entire $4 Walmart pillow inside my (sealed) DIY guitar cab; I heard an audible improvement in sound (smoother midrange) when I chucked it in and re-sealed the box.
I imagine loosely filling your enclosure with Walmart pillow-stuffing would be a quick and easy way to find out if some acoustic damping would benefit your DIY line array.
Nice, that will get around the 140 Hz low-end-limit problem of the 2.5" drivers.
Good luck with your design - I hope you figure it out, get it right, and end up loving your speakers.
-Gnobuddy
Just to provide a little reset here though. The "eye on the prize" so to speak. Google the BOSE L1 Model 1S system with B1 bass. It can be had for about $1700. The amplifiers? The "Bass Box"? The EQ? No magic in any of that, and I've got all i need there anyway. But that column? Whenever I've heard one of these systems in a club, I can immediately see the benefit to every listener in the room, and the musicians (not to mention the likelihood of being invited to return to perform again, because of the great sound!). So I understand that its only "magic" because its a new approach to me, and its REAL... it WORKS! And... the fact that others you've heard sound like crap further reminds me that I need to better understand WHY this one is a winner, to maximize my chance that mine will be too!
Considering what musicians get paid around here (trust me... not a kings ransom!
), there are some good reason why these musicians are buying these up. Now what I need is for someone to saw one open for me, and tell me what goes on in there! As you've pointed out, there are a LOT of reasons why it should NOT work. Why it does is not magic, but there are still some mysteries here to be uncovered! 
Nice set of measurements!
I do wonder, though, if floor-to-ceiling arrays still "count" here, though - you've got reflections at the floor and ceiling to extend the array, as discussed by Keele, but for gigs, I'm not so sure that applies.
Peter Pan, I'm in (almost) the same boat as you. I wanted to DIY a Bose L1 killer. Something that'd go louder, cleaner, in a similar package.
The variation in HF with distance is something you'll need to deal with for a line array in free space. I stopped the bottom 9 drivers from doing treble, and it pretty much works. Its now just a wide-dispersion point source with some vertical directivity.
Chris
If we look at ra7s measurements:
The 12 ft and 6 ft measurements are quite alike. The 3 ft shows more combing, from being too close to the array.
To our OP I'd like to say: pick a driver you like, that has a clean impulse by itself. preferably with a slightly rising top end. Picking the right driver is crucial i.m.h.o.
That's why the TC9 is popular for line arrays (not because it's cheap). It's clean, even in off axis measurements as Art (weltersys) demonstrated in this graph:
Even though some people point out it's only a TV speaker, it's a very clean and capable full range driver. One of few that pulls this off. The low price is a bonus.
Be sure your preferred driver is up for that task. The newer TC drivers are different and have no real family resemblance to the TC9 line array favourite.
We've had someone try an array with one of the newer drivers but it didn't work out quite as well. http://www.diyaudio.com/forums/multi-way/288784-tc6wd-pe-buyout-line-array-lessons-learned.html
Another concept definitely worth looking at is the SBH series speaker from Danley. It looks like a line array but it's actually a point source due to the clever use of the Paraline lens shaping the wave front. This is a great concept for gigs. But much harder to DIY. Not many have succeeded to DIY a Paraline yet. But this will probably kill any variant made by Bose.
The arrays as a concept that ra7, Halair, OPC and myself (and probably a few others) use is more of a home use floor to ceiling concept.
The 12 ft and 6 ft measurements are quite alike. The 3 ft shows more combing, from being too close to the array.
To our OP I'd like to say: pick a driver you like, that has a clean impulse by itself. preferably with a slightly rising top end. Picking the right driver is crucial i.m.h.o.
That's why the TC9 is popular for line arrays (not because it's cheap). It's clean, even in off axis measurements as Art (weltersys) demonstrated in this graph:
Even though some people point out it's only a TV speaker, it's a very clean and capable full range driver. One of few that pulls this off. The low price is a bonus.
Be sure your preferred driver is up for that task. The newer TC drivers are different and have no real family resemblance to the TC9 line array favourite.
We've had someone try an array with one of the newer drivers but it didn't work out quite as well. http://www.diyaudio.com/forums/multi-way/288784-tc6wd-pe-buyout-line-array-lessons-learned.html
Another concept definitely worth looking at is the SBH series speaker from Danley. It looks like a line array but it's actually a point source due to the clever use of the Paraline lens shaping the wave front. This is a great concept for gigs. But much harder to DIY. Not many have succeeded to DIY a Paraline yet. But this will probably kill any variant made by Bose.
The arrays as a concept that ra7, Halair, OPC and myself (and probably a few others) use is more of a home use floor to ceiling concept.
To our OP I'd like to say: pick a driver you like, that has a clean impulse by itself. preferably with a slightly rising top end. Picking the right driver is crucial i.m.h.o.
That's why the TC9 is popular for line arrays (not because it's cheap). It's clean, even in off axis measurements as Art (weltersys) demonstrated in this graph:
Two reasons I picked (tentatively anyway) the Vifa/ tymphany tc7fd00 2-1/2" speaker was first because of its impressive frequency response graph (just look at the graph in this short spec. I'd post the image directly, but it seems the forum requires a URL and won't let me directly upload an image)
http://www.parts-express.com/pedocs/specs/264-1144--tymphany-tc7fd00-04-spec-sheet.pdf
My second reason is that having used these in a very unusual speaker project (totally different from an array) and had surprisingly good results, I have a lot of them. They aren't cheap either IMHO. The only complaint I'd have with this speaker, based on the specs alone, is the sensitivity (which I assume equates to a standard SPL measurement). But it does stand to reason that if you want flatter response its going to cost something, and in this case the cost seems to be that more power will be required. But that won't be an issue. Any thoughts on this speaker?
Yes, I have been a little frustrated and maybe surprised there is so much emphasis on home use, and floor to ceiling arrays, when the people raving most about these are musicians and their surprised audiences. I'm going to continue to read all I can from various DIY experiences, and in the end it seems I'll have to go with my gut with various construction options and give it a try. I DO wish, especially considering the price and success of the BOSE L1, there was more info on exactly what's inside the box. What angles (left and right) to they mount their small speakers? Are there baffles or blocks on the inside to break up standing waves? Are they packed with foam or is it one big empty chamber? How exactly do they wire the speakers? (louder on top? Louder in the center?, equal all the way?) There's just very little hard information out there. There ARE a LOT of people telling how BAD bose designs are. But they are missing the fact that THIS product, used in their intended situations, is a total winner sound wise.
Last edited:
Genuinely curious about this Bose design.
My experience with Bose (home) has been that it's bought by people with deep pockets and shallow tastes.
Would like to hear some pro audio guys chime in? Again just curious. Find is hard to believe that given the muck (crap) they churn out to the hi fi buyer they somehow produce quality in the PA arena?
Totally willing to be convinced otherwise
My experience with Bose (home) has been that it's bought by people with deep pockets and shallow tastes.
Would like to hear some pro audio guys chime in? Again just curious. Find is hard to believe that given the muck (crap) they churn out to the hi fi buyer they somehow produce quality in the PA arena?
Totally willing to be convinced otherwise
Genuinely curious about this Bose design.
My experience with Bose (home) has been that it's bought by people with deep pockets and shallow tastes.
Would like to hear some pro audio guys chime in? Again just curious. Find is hard to believe that given the muck (crap) they churn out to the hi fi buyer they somehow produce quality in the PA arena?
Totally willing to be convinced otherwise
People do indeed like the Bose L1 and others like it from other companies, because they are easy to setup, and do a very good job for small to medium sized venues. Easy to carry is nice too.
It is aimed at good output, a reasonable frequency range, and portability. In those areas, they succeed.
It is not aimed at the hifi crowd, looking for meticulous details in the reproduction, nor spatial imaging and the likes. The users will pan the sounds either right or left (if you have two units) in a mixer and that's about it.
So, yes, they are well liked in the area they fit in, but I wouldn't use them for "critical" listening... whatever your definition of critical listening is.
So just a quick update! First, thanks to all those who talked me into setting up some real software tests. I've since installed the free RightMark audio analyzer and also the ARTA packages, both of which work with my sound card. Fortunately I have a couple of very good Samson condenser mics designed for recording acoustic guitar, and it turns out their curves are pretty darn flat, apparently from DC to daylight (J/K).
Anyway, I did an initial assembly and ear test first. My array ended up with 15 of those 2-1/2" VIFA speakers, and I included 4 multi-point terminal mounting strips inside, to make wiring experimenting with different combinations of power weighting easy. To start I just paralleled 3 groups of 5 (4 ohms each) in series, for a total "demonic" impedance of 6.66 ohms.
Before any measurements, I tried my array by ear, in its expected environment. That is, the array sits 2 feet up, on top of my bass box, a ported bass reflex speaker. The ear tests were great!! I did have to turn up the high treble end on my mixer about 4 dB since no tweeters are in the mix. But the sound with acoustic guitar and singing were awesome. Using a wireless link I was able to move about a 20 foot room and feel i'd achieved the first goal: clear articulate sound over at least a 120 degree range. FYI, 14 of the speakers are alternately shaded left and right, by 15 degrees, and the 15th speaker on top points straight out.
So now on to the actual measurement testing, of the array by itself (no bass box). As this was intended as a midrange and upper mid array, I wasn't too disappointed to see response drop off pretty quickly below 200 hz. I was happy to see robust response out to about 15K. But I was disappointed to see some obvious mystery resonance peaks at 1K, 1.5K, and another at about 8.5K. But then, after plugging these into a calculator I got wavelengths of 13" , 9", and 4". Then i remembered... I have 4 structural braces across the inside extending from flush with the back (to receive screws), and extending about an inch less then going all the way to the front baffle. Well this makes perfect sense! 3 of the support braces are 13" apart, the last one is 9, and inner width (and depth) of the array box itself is about 4".
This may seem discouraging, but actually is very satisfying to be able to mentally map each "bump" on a response curve to an actual structure. Especially since by ear and with real guitar and vocals, the array is already a success.
So the next think I need to do is some experimenting to combat those resonances. Obviously putting dense soft craft from on all inner surfaces, along with tightly stuffing the box with teddy bear pillow foam was not enough. So what I'm thinking of doing now, is I'm going to cut some light wood into 3-1/2 " strips, maybe 10 inches long, and mount them internally at angles. spanning from one inner wall across one side to the other. My plan here is make an apparently simple resonance situation into a very complex situation, with multiple broad angles everywhere. The hope will be to cause nearly infinite resonance points, so that they all become irrelevant. I suppose I'll have to cover these inner baffles with more of that soft foam, and then push back all the pillow stuffing. Maybe something even denser like pieces of an old blanket. I do understand that I could also opt for a series of resonant LC+R circuits to force the response flat. But I think it makes sense to do everything mechanically possible first. Good build before EQ, right?
In any case, I'll keep up the thread with my progress, and eventually post some photos and curves, once I decide I've done all I realistically can. I'll probably post all my assembly photos on my own website and share a link.
Anyway, I did an initial assembly and ear test first. My array ended up with 15 of those 2-1/2" VIFA speakers, and I included 4 multi-point terminal mounting strips inside, to make wiring experimenting with different combinations of power weighting easy. To start I just paralleled 3 groups of 5 (4 ohms each) in series, for a total "demonic" impedance of 6.66 ohms.
Before any measurements, I tried my array by ear, in its expected environment. That is, the array sits 2 feet up, on top of my bass box, a ported bass reflex speaker. The ear tests were great!! I did have to turn up the high treble end on my mixer about 4 dB since no tweeters are in the mix. But the sound with acoustic guitar and singing were awesome. Using a wireless link I was able to move about a 20 foot room and feel i'd achieved the first goal: clear articulate sound over at least a 120 degree range. FYI, 14 of the speakers are alternately shaded left and right, by 15 degrees, and the 15th speaker on top points straight out.
So now on to the actual measurement testing, of the array by itself (no bass box). As this was intended as a midrange and upper mid array, I wasn't too disappointed to see response drop off pretty quickly below 200 hz. I was happy to see robust response out to about 15K. But I was disappointed to see some obvious mystery resonance peaks at 1K, 1.5K, and another at about 8.5K. But then, after plugging these into a calculator I got wavelengths of 13" , 9", and 4". Then i remembered... I have 4 structural braces across the inside extending from flush with the back (to receive screws), and extending about an inch less then going all the way to the front baffle. Well this makes perfect sense! 3 of the support braces are 13" apart, the last one is 9, and inner width (and depth) of the array box itself is about 4".
This may seem discouraging, but actually is very satisfying to be able to mentally map each "bump" on a response curve to an actual structure. Especially since by ear and with real guitar and vocals, the array is already a success.
So the next think I need to do is some experimenting to combat those resonances. Obviously putting dense soft craft from on all inner surfaces, along with tightly stuffing the box with teddy bear pillow foam was not enough. So what I'm thinking of doing now, is I'm going to cut some light wood into 3-1/2 " strips, maybe 10 inches long, and mount them internally at angles. spanning from one inner wall across one side to the other. My plan here is make an apparently simple resonance situation into a very complex situation, with multiple broad angles everywhere. The hope will be to cause nearly infinite resonance points, so that they all become irrelevant. I suppose I'll have to cover these inner baffles with more of that soft foam, and then push back all the pillow stuffing. Maybe something even denser like pieces of an old blanket. I do understand that I could also opt for a series of resonant LC+R circuits to force the response flat. But I think it makes sense to do everything mechanically possible first. Good build before EQ, right?
In any case, I'll keep up the thread with my progress, and eventually post some photos and curves, once I decide I've done all I realistically can. I'll probably post all my assembly photos on my own website and share a link.
Nice work! Based on what I've seen of Bose products, I suspect you've already done more R&D and engineering than Bose ever did.
I'm a bit shocked that even stuffing the entire enclosure full of foam still wasn't enough to control those pipe resonances. Clearly, long, skinny, pipe-like enclosures are tough beasts to tame! I guess if it is a duck, it's not terribly surprising that it wants to quack like a duck!
-Gnobuddy
Success! So I think I may be on to something with that idea of adding angular baffles inside a line array. While I won't say its curves are flat as pancakes, the resonances did go away after that addition! Again, this array is intended for use with a separate bass box, so its normal that by itself it drops off on the low end. But it sounds great, and only needs a little EQ on the high end now. (Ignore the green trace... thats supposed to be THD). Also, its really amazing how it sounds too! You can stand anywhere in the room, and if you close your eyes and turn, you can't pinpoint where the sound is coming from. Its almost like wearing headphones in that sense! But anyway, thanks again everyone. I'll add some photos of the finished work as soon as I finish the front grill.
Attachments
Of course the angles I was talking about in my last post were about those inner baffles. But for the drivers themselves, 7 of them are pointed 15 degrees left, 7are pointed 15 degrees right (alternating of course), and the topmost speaker (#15) pointed straight out. This was based on what the BOSE speaker angles look like to me, as well as the graphs for the drivers I picked (Those Vifa/Peerless speakers). The graphs provided in the data sheet for those speakers show the flattest SPL curves facing head on, as well as additional curves 30 and 60 degrees out. The speakers are great right up to the end of the audible range head on, but definitely show significant loss in the highs as you get up past 30 degrees off center. I reasoned that the only way I was going to have anything close to consistent response over a fairly wide listening angle was to split the difference, and have at least ONE speaker facing forward. So listening to the array head on, at least one speaker will point right at you, and the others all seem to blend well. I still have to take more measurements at different angles. But to me, as a musician, it sounds great to my ears, and its literally weird the way walking around the room hardly changes anything. That's probably at least partly due to the fact that the room itself is pretty hot. I really need to try it set up outdoors. But wow! When I first jacked in and heard it, I must have spent 3 hours with a wireless link connected to my acoustic/electric guitar playing everything I could think of while walking around the room. This, hands down, is probably the most successful speaker I've ever made. I'll post some photos soon!
I promise! And after that, I'm going to put up a page on my own (web site) explaining enough details for a seasoned builder to duplicate what I did. Not a granular step by step (drill this screw hole at this four digit measurement), but definitely enough. Studying the topic of line arrays, one thing I've learned is that a lot of things can go badly wrong, and whether by ingenuity or dumb luck mine worked. So I would really have liked to see these details before I started!
By the way... I'd have posted some pics already if not for the fact that 2 days ago I discovered my cat had shredded the only USB connector wire that actually fits my stupid cybershot camera (next camera I get is going to have wi-fi!!!). All the pics i took so far are on it, and my i-phone-4 camera is total crap! New cable is on its way from e-bay. Should be here within two days!
Last edited:
Still waiting for my stupid replacement camera cord, but I will. My response curves are not as flat as yours, but in my case I wasn't expecting or planning for much of anything below 200hz, since I'm using a separate more traditional bass box. So many things I've learned from this and many other discussions, but a big scary one was the realization that designing one of these for a performance venue has very different rules than designing for a room. In a way I've been very lucky because there is very limited information when it comes to performance versions and I realized I was going to have to make my best guesses about a lot of details. Mine isn't even a floor to ceiling array... just 4 feet end to end, mounted now on a pole about 2 feet up.
By the way, I hope I haven't hijacked your thread posting here, or adding some pictures soon.
As promised, here are some random photos I took along the build of this speaker array. I'll just upload 1 to the forum, but I'm also going to provide an open link to a directory on a domain I own, where you can see whatever photos I took. When I have time, I'll organize that collection into an actual web page with measurements and such, for anyone that may want to somewhat duplicate what I did. For now at least you can look through the photos. Toward the end you see those angular baffles I added, which pretty much eliminated the standing wave resonances I described a few messages back. Thanks again to everyone that offered a lot of input here. You've helped me add a kick-a$$ piece of equipment to my performance gear, as good (to my ears) as that Bose L1, and I barely paid $250 for all materials (including drivers) to build the array. Here's one selected image of the near finished product.
And here's a link to a directory with over 30 images of various construction details. Hopefully this will work with the forum editor
http://elfintechnologies.com/pubImages/spkrArray/
An externally hosted image should be here but it was not working when we last tested it.
And here's a link to a directory with over 30 images of various construction details. Hopefully this will work with the forum editor
http://elfintechnologies.com/pubImages/spkrArray/
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.