horn top splaying |
Post Reply | Page 12> |
Author | |
fourway hornloaded
Registered User Joined: 16 March 2014 Location: Hamburg Status: Offline Points: 60 |
Post Options
Thanks(0)
Posted: 16 March 2014 at 1:20pm |
Hi all,
I've just registered to this forum, so I'll start with a few words about my involvement in this hobby. I've been reading on here since eight years or so. Back then, me and a couple of friends on a tight budget had some great fun building a punisher from MDF, later a pair of MBH115s. Around then I discovered hornresp and got acces to a nice workshop. I chewed around on expanding the usable range of the HD15 design and finally built a pair 15" BPH6's with a ported back chamber, but shaped like a W-bin. Loud and punchy but lacking low-end, so I redesigned them into a 18" version using PD186 and built 6 pieces. I built a pair of Limmer P3 in a horizontal layout to go on top of them. Well that works very well! I think the P3 is generally considered a good design and I can only recommend it. But I am struck by this fever some of you might kow ;-) so i'm thinking of doubling up the amount of BPH6 bassbins to 6 per side. And then, correct me if I'm wrong, I think I'm entering a stage where one hornloaded top per side is not going to match the bass anymore. Generally speaking, when using multiple tops per side in a horizontal row, you'd say you need narrow dispersion, about 30-40 degrees. Three pcs of 40 degree tops a side would be perfect, 120 degrees in total. Wouldn't you? Reading around about CD-Horns and horn dispersion in general, tinkering with horn-resp a bit, I'm trying to get an idea of what can be done. I'm an RF engineer and I'm familiar with the theory. So now to my question of how to properly splay tops: In the lower-mid range up to ~500 Hz, normal-sized speakers don't beam, they just couple. A horizontal row of three tops has wavelength-like dimensions in this frequency range so the resulting dispersion will have cancellation zeros off-axis. For example a 150cm wide horizontal row should have its first zero at around 100 Hz at 90 degrees of axis. Entering the upper mid range, >500Hz, where speakers do beam. As I understand it, you'd be trying to splay the tops so that each -6dB off-axis response overlaps with the next. Splay angle then equals coverage angle. If I recall correctly, large Funktion One (or Turbo) speaker clusters are set up so, that each person in the audience has only one top bin pointed at him. For a 3-top 120-degree cluster you'd then need three 40-degree tops a side. Now there don't seem to be many of those 40 degree top bins. I am confused. For example, a Turbosound TSE-111 has 70 degrees of horizontal coverage. If you'd try to splay those bins, having each bin cover 70 degrees of the audience, you'd have 70 degrees of splay angle between them. Ugly! More over, you could not use more than two of those tops per side without either breaking the 'rule' or having more than 180 degrees of coverage. Maybe that 'rule' (coverage=splay angle) is not correct? TMS2 and TMS3, that are based on the same mid-horn, are usually splayed only 10 degrees or so. In the TSE-111 datasheet, there's a sketch of the flying hardware, showing ~35 degrees horizontal splay. Another example is the Limmer 308, which has a well-documented coverage of 60 degrees. You can find the coverage graphs on a resellers' website. (Other manufacturers should take an example and publish those colored graphs as well). Anyhow, you could only use two Limmer 308s a side to get 120 degree coverage, as I understand it. Where am I going wrong? No-one splays top bins by 60 degrees. Is it generally accepted to have cancellation and comb-filtering all over the place? How do you guys splay? I'm hoping on a nice and nerdy discussion! Thanks in advance and have a nice day, Robert |
|
fourway hornloaded
Registered User Joined: 16 March 2014 Location: Hamburg Status: Offline Points: 60 |
Post Options
Thanks(0)
|
Maybe my first post was a little too long. I will try to summarize:
When using multiple tops per side, * Which horizontal dispersion do such tops need to have? * Which splay angle do you tend to use? I've posted some pictures of my system with the Limmer P3 tops in the 'show off your sound system II' thread. Robert |
|
dylan-penguinmedia
Old Croc Joined: 14 April 2011 Location: Brighton Status: Offline Points: 4575 |
Post Options
Thanks(0)
|
This will vary with the room you are setting up in, and the tops you are using - there isn't a definitive right answer to this one really! If you have a wide, shallow room, you're likely to need wider dispersion than if you have a long, narrow room.... |
|
burningbush
Old Croc Joined: 25 May 2009 Location: Pictland Status: Offline Points: 5897 |
Post Options
Thanks(0)
|
Set an approximate splay, then walk perpendicular to the stack listening for an even spl. If you hear NOTICEABLE drop outs or peaks then start widening or narrowing until you get the best compromise. It is all a compromise, and you have to ask who will notice?
|
|
music is the message
|
|
bitSmasher
Old Croc Joined: 23 June 2012 Location: Melbourne Status: Offline Points: 2295 |
Post Options
Thanks(0)
|
|
|
fourway hornloaded
Registered User Joined: 16 March 2014 Location: Hamburg Status: Offline Points: 60 |
Post Options
Thanks(0)
|
Thanks for the replies. I've heard some multiple-top systems that had the entire mid-range fade away when walking past them from left to right.
So I reckon once you've got a bunch of midhi cabs, you're going to have to try around a bit with the set-up. And I grasp the shallow-room thing. But I'm more on the theoretical side, still. Now how would you reason when designing a 3-top-per-side sound system for outdoor usage. If you're aiming at larger events that simply can't be done with one top per side. - Would you build a bunch of 'ordinary' horn tops with 60 degree dispersion and give it a go? Like MT-121, Limmer-308, Xtro, p0rn-horn. - Or would you say: don't go there, it's going to be a mess. Because there aren't that many 40 degree horns about ‼ Actually, the only thing I found is a 2" JBL 2386. Has anyone stumbled across the same problem? Regards, Robert |
|
Teunos
Old Croc Joined: 23 November 2008 Location: The Netherlands Status: Offline Points: 1799 |
Post Options
Thanks(0)
|
And the 2386 is not even anywhere near constant directivity. When i saw your post i was just about to post a topic about large format 2'' horns.
what you state is entirely correct, you will never get perfect coupling from multiple drivers working side by side, you will always have interference, which you as an RF engineer will understand. Drivers and drivers never have a perfectly defined dispersion angle unless they are infinitely large. Using multiple drivers in the same band is always a compromise when wavelengths become small compared to distances between drivers. When doing the design properly, you can however get around the bumps. The biggest problem is finding components that behave like a constant directivity source. This means that the lower the frequency, the longer the horn and the bigger the radiating area has to be. Like you indicate, above 500Hz you will start to notice comb filtering. when using a GOOD CD horn, the best thing you can do is measure the dispersion of the horn yourself, and then measure the dispersion of the woofer below it. From this, choose a crossover point where both dispersion and (-6dB) dispersion gradient match. The latter being equally important because otherwise you will still get suckout off-axis. If the directivity gradients are nowhere reasonably equal to each other where dispersion of woofer and horn match, you have chosen the wrong size driver/horn combo. (e.g. 15'' woofer with 1'' driver on small 90*60horn or 10'' with huge 60*40 horn) From there on eq every driver to be flat within at least half an octave but preferably a full octave from the crossover point, or at least so that there are no big discontinuities in the phase or frequency response. Then set up the crossover at the desired point. Many people forget this step, and only measure on axis which causes the system to not behave well off-axis. When multiple cabs are needed per side, this process (for me at least) is an absolute must. My preference for multiple tops per side, and i have both, is either a large format single 12 or preferably double 12 horn with a large format horn above, or either a single/double 15'' BR with a large format horn. 12'' BR plus horn never really attracted me for multiple cabs per side, although i am fully aware of the benefits of a 12 vs a 15 doing mid duty. Chosen the right compression driver however, mids can still sound great from a comp driver crossed low.
|
|
Best regards,
Teun. |
|
Teunos
Old Croc Joined: 23 November 2008 Location: The Netherlands Status: Offline Points: 1799 |
Post Options
Thanks(0)
|
BTW, at the moment i am using BMS 4590 2'' comp drivers on JBL 2385 horns with double 12'' horn loaded mids below, and i truly feel the JBL horns have become the weakest part of the system as they are not the best when it comes to controlling directivity. It is about time a manufacturer stepped up with a large format 2'' CD horn that can actually be crossed at 600Hz without compromise, and please without using a diffraction type throat.
I would love to see an enlarged version of the XT1464, which performs excellent combined with the NSD1480N right down to 900Hz crossed to a 15'' with a light cone.
A bit like the tannoy VQ64, but then with proper mouth termination. Or something based on the build quality of stereolab's horns but then CD instead ot tractrix. Edited by Teunos - 17 March 2014 at 10:10pm |
|
Best regards,
Teun. |
|
fourway hornloaded
Registered User Joined: 16 March 2014 Location: Hamburg Status: Offline Points: 60 |
Post Options
Thanks(0)
|
Thanks for explaining the crossover/beamwidht thing Teun! You seem to
work with quite a few different mid-hi bins. I was thinking along the
fully hornloaded route, although double-15" top bins like you mention,
crossed my mind as well. Though it's clear to me that
larger bins are the ones that have narrow dispersion, I am just so
tempted to build something smaller. Take for example Funktion One Res3,
if you'd leave out the 18" horn, you'd have a nice little 65cm x 65cm
bin doing 180 upwards with 50 degrees of horizontal coverage. I'm not
sure what to believe though. The double-12" horns you were talking
about, for example, do not seem to reach less than 60 degrees below 800
Hz, even though they are large:
http://www.loudspeakers.de/daten/limmer/lim-302.jpg http://lsv-achenbach.de/daten/limmer.htm As for 2" cd horns without diffraction slot: did you see Limmer 243? I don't want to sound like a complete limmer fan here, but his dispersion plots look nice. I had a look at the JBL 2385 after reading your post... and got the point. 'so much for constant directivity'. Your double-12 seem to be quite brutal if they outplay such expensive compression drivers. I reckon you're in quite a different ball park then. Myself, I'd like to play around with small-dispersion paper-cone mid-horns in the future. After reading Volvotreter, Keele, Geddes and especially Peavy's 'quadratic-throath' paper, I'm actually not put down. We all know Martin, Turbo, F1, Föhn, Limmer and even RCF have been very succesfull with paper-cone mids. Now recently if you have to believe Peavy and Geddes, the whole horn-business seems to steer away from cd-horns, knowing you aren't going to get constant directivity anyway, and turn towards conical horns, with maybe a little curvy bend at the mouth to avoid too large a mismatch. Conical horns have narrower dispersion than similarly sized exponentials. Especially in a multi-way system you should be able to cut off such a horn before dispersion gets too narrow, or at least before the first transversal reflections in the mouth occur. Conical is of course very DIY-friendly. Well, that doesn't solve the problem of finding a narrow horn for your high-frequency compression driver, of course. Unless you'd want to make that yourself too. Right now I'm figuring if this high-fi conical thingie, I mean oblate spheroid and the like, are going to work equally well with narrower dispersion. Because neither Peavy nor the hifi guys seem to do anything narrower than 90 degrees. A narrow conical horn might very well turn out to sound like a megaphone... because it actually 'is' one. What is with elliptical, what can it do? The XT 1464 you mention is elliptical. Hornresp doesn't simulate it. On the forums, they're all into that type of horn. |
|
fourway hornloaded
Registered User Joined: 16 March 2014 Location: Hamburg Status: Offline Points: 60 |
Post Options
Thanks(0)
|
Edit: the double-12 horn does 60 degrees from ~450 upwards rather than 800Hz.
|
|
AndyWave
Registered User Joined: 30 March 2008 Location: Finland Status: Offline Points: 380 |
Post Options
Thanks(0)
|
Proper mouth termination is the key not having megaphone sound.
I'm using quadratic throat 80x40 horn with 630mm x 330mm mouth and BMS4594, horn is terminated with about 75mm radius to the outside air, xo at 900Hz. Horn is about 270mm deep Two 670mm deep straight mid-bass horns below with 18 sound 12ND930. Whole cabinet is 800mm deep. Result is 110dB/W above 150Hz after eq:ing the peaks down Btw horn response doesn't take account if your horn's axial geometry is a square, ellipse or round. Or bent like bass horns. Only expansion rate counts here. Ellipse is best (most hifi, at least for me) compromise to have H and V in different angles. My small pa uses Faital FH144 and LTH142 combo, really sweet sound compared to Beyma 380M+TD164. And Teun, that 600Hz horn must be quite huge to give acceptable loading down to 300Hz:ish. Not that you wouldn't know that already but is it practical... edit:Sorry got bit carried OT I would try to make the high horns with 40 horizontal first and modifying them so long they would comb as little as possible, after that mate them with normal 2x 12" BR low mid section with cabinet so narrow as possible to achieve good coupling up to about 800Hz. This would dictate the cabinet width somewhere 360mm. EV phoenix 2122 as example. Form follows function. my 2p Edited by AndyWave - 04 April 2014 at 7:42am |
|
torturing electrons since ......
|
|
snowflake
Old Croc Joined: 29 December 2004 Location: Bristol Status: Offline Points: 3122 |
Post Options
Thanks(0)
|
theory says conical horns splayed at the same angle as the horn wall will give the most seamless array. if you vertically raise adjacent boxes as you go across the array you can arrange the horn walls of adjacent boxes to be exactly above each other. the hf is controlled by the horns and no listener can ever be in line with more than one horn mouth. the lf is omnidirectional from a single box but the array is large in both horizontal and vertical dimension so gives some directivity and avoids the 'pattern flip' problem. so your array would look like this:
[] [] [] or this would be nearly as good and could be stacked rather than flown: [] [] [] just bought the wood to build enough boxes to test this in practice so will let you know in a month or so ;) |
|
Post Reply | Page 12> |
Tweet |
Forum Jump | Forum Permissions You cannot post new topics in this forum You cannot reply to topics in this forum You cannot delete your posts in this forum You cannot edit your posts in this forum You cannot create polls in this forum You cannot vote in polls in this forum |