|
|
Post by chrisbee on Oct 23, 2009 3:42:58 GMT -7
I have been itching to have a proper discussion regarding the seemingly obvious, acoustic deficiencies of manifolds relative to arrays. So here's your chance to poke holes in my fuzzy assumptions. Absolutely no criticism is intended towards manifolds. I have used two of them myself following the trials and tribulations of my original line array.
We probably all agree that a driver array housed in a massive wall is the perfect IB. The drivers suffer no loss of output due to flex in the baffle. The drivers see no near reflections from either the enclosure or the AV room. Being acoustically transparent we like our cones to see only fresh air. We don't want any nasty reflections passing through the cones to spoil our sound quality. Or returning (out of phase) pressure waves damping our cone movements at selective frequencies.
The larger the enclosure and AV room the further away are the reflective boundaries. Hopefully the returning sound waves are all the weaker for greater distance. (the inverse square law may not apply perfectly indoors but it's close enough)
An IB using the great outdoors for an enclosure is as good as it gets. As long as nothing leans against the backs of the drivers there should be no reflections. There would still be some residual reflections from the ground but these could be weakened by distance, ground slope and/or dense planting.
Then there's the real world the rest of us live in. We don't live miles from our neighbours and our inside walls aren't built like a concrete dam. So we choose a manifold to cancel structural vibrations by opposing the drivers. We also use a finite, indoor enclosure to limit the amount of bass escaping to our trigger-happy neighbours.
A manifold is a sealed box which is wide open on one side. It is a local folding of our, usually flat, IB baffle. The open side reduces the number of solid surfaces, surrounding our drivers, from 4 to only 3. Does this mean a 1/4 reduction in colouration?
When the drivers are facing each other across a manifold there is a constantly changing (dynamic) relationship in the opposing surface as the cones move towards and away from each other in unison. They see a mirror image of themselves. The cones are also acoustically transparent so that's slightly in our favour. But then so are other subwoofer designs using opposed drivers provided there aren't too many solid, bracing baffles between them.
What about the solid top, bottom and back of the manifold? Aren't they all very close to our drivers? We wouldn't dream of facing our drivers to a solid wall from such a short distance. So why do we insist on stuffing our drivers into this small box? For compactness and convenience?
An array is usually far more compact because the depth of the drivers can be lost in the wall. If only the array wouldn't throw away it's advantages to structural vibration! In a stud wall an array may produce more colouration from vibrating surfaces than the drivers output themselves. A wall has a huge area and the slightest vibration produces huge output. Or huge damping effects. Whether the output or selective cancellation from the wall would be in phase with our drivers at all frequencies is quite another matter. Would that we could use a whole wall as one huge IB driver! Cone excursion could be reduced to a fraction of a millimetre at deafening levels deep into the infrasonics! But I digress...
Why do manifolds work so well? Because we use a low crossover to avoid listening to colouration from the close reflections? Well, so does everybody else! With their little boxes and big boxes and cylinders regardless of whether they are ported or sealed.
The drivers in a sub usually see an open space on one side whatever tortures they might suffer on the other side of the cone. Only downfiring subs seem to shut themselves right in on both sides but they are still incredibly popular.
Is having one open side to our manifolds really enough to overcome the nearness of all the other surfaces of our box? Or is it simply enough that our drivers are still working in an IB?
|
|
|
|
Post by ThomasW on Oct 23, 2009 8:11:08 GMT -7
As has been posted numerous times, a manifold functions like a compression driver. No doubt the use of a manifold imparts some coloration to the sound. The audibility of the coloration is questionable given the extremely low frequencies in the passband. The success of Tom Danley's Unity horn design tells us it's not that problematic having drivers fire into a small space. In this design 5" cone drivers fire through a small hole into the main horn. The passband for these drivers starts at ~350Hz  Designs like the Synergy horn have even larger drivers firing through smallish openings  And before it's asked, yes there's potential benefit from using a manifold with angled sides. I'll leave it to someone more motivated than I to compare/contrast straight vs angle sided designs. All the above aside, in theory a perfectly braced line array should provide 'better' performance, all other things being equal. |
|
|
|
Post by jimcant on Oct 23, 2009 18:43:24 GMT -7
Chris, I can tell from reading your latest blog entry, that you are up to something, and this thread further fuels my suspicions  What happened to "if it ain't broke, don't fix it"? Perhaps what we have here is "keep on making improvements until it is broken" ;D Good thread though  I always enjoy learning more about this topic. I think I have at least one more IB incarnation left in me, and it will be a pair of two driver vertical arrays on my front wall. In the mean time I will ponder how much concrete I need to make them immovable this time  Cheers, Jim. |
|
|
|
Post by chrisbee on Oct 24, 2009 0:01:25 GMT -7
Thankyou, both, for your responses. Though I can't say I feel more confident in my understanding of the effects of putting our drivers in a small box between two large spaces. What I am really trying to get at is: Why don't we suffer from boxy colouration? My manifold IBs don't suffer from the very obvious colouration of my previous subwoofers. I know from personal experience that a sealed spherical enclosure sounds amazingly open and "clean" compared with the box of the same volume which originally housed the same drivers. Spheres are out of the question for the enclosure itself but may offer new ideas for a manifold if a suitable "donor" hemisphere can be found. Jim I am still considering a major redistribution of wealth. ;D I have a wealth of drivers and they are not enjoying accommodation which matches their expectations. Their enclosure is simply too small which limits their ULF extension. It sounds amazing and goes incredibly loud without visible excursion. But... I thought it would be easy to find large sheets of thin, flexible rubber to allow me to experiment with membranes over manifold mouths. The idea was to trap still air and block indoor humidity from reaching the cold drivers. Hopefully without affecting sound quality. Only experiment would tell whether this idea would work in practice. I hoped it would allow much greater freedom of location in my cold winter climate. I desperately need to overcome the thermal loss problems before I dare to use my other enclosure options in the remaining roof space. Perhaps I need to use the cones themselves in isobaric configuration to keep the heat indoors? One of the potential sites is in a dormer ceiling right above where I sit at my computer. I have enough problems with windows drivers without having to mop up condensation dripping from my IB drivers. I'm fairly sure it's not what they mean by "water cooling". |
|
|
|
Post by weverb on Nov 28, 2009 9:04:53 GMT -7
And before it's asked, yes there's potential benefit from using a manifold with angled sides. I'll leave it to someone more motivated than I to compare/contrast straight vs angle sided designs. I love how Thomas teases us with hints of information. Chris, I have read your blog and many posts and find them very interesting and love how you keep testing theories. ;D I have a few more questions to throw at you and Thomas along this topic. 1. Does manifold orientation also have any effect? Is there a difference between manifolds mounted above that fire down at the floor versus manifolds mounted in the wall that fire directly at the listening position? 2. Is there a difference if the manifold only has two drivers versus four? Now you just have a solid top. 3. Don't we need to keep the room in mind when comparing these? Jim is the only one I have seen that has tried an array and manifold in the same room. Since he moved everything around, it is still not an apple to apple comparison since the room probably has changed how things reacted.  |
|
|
|
Post by jimcant on Nov 28, 2009 20:04:43 GMT -7
weverb, Glad you are keeping this thread alive I too would like to know the answers to your questions. With four drivers in a manifold, as opposed to two, there would be a greater degree of coupling. With regard to #3 though, I am certainly not the first to have an array and a manifold in the same room. A number of builders, including Chris have changed their arrays to manifolds. Did you perhaps mean, at the same time, in which case I do not know of anyone else silly enough to try that |
|
|
|
Post by chrisbee on Nov 29, 2009 1:33:58 GMT -7
Thanks
Ironically, Jim's shelf suggestion may actually have helped to decouple my two different driver sets. Or the whole thing was simply too flexible despite the steel studding braces. I made abbreviated shelves for the upper and lower sets because they didn't need acoustic decoupling.
My baffle wall and door remain my Achilles heel so nothing I do will make much difference now. It still sounds awesome on live programme material so I don't care any more.
I promised myself to make an OB for my old SVS PCI driver on the right side but it would have to be folded into a deep sided box to fit my limited stage space. Only the unreachable trough at 150-200Hz spoils my response now. This is the chest thumping caveman area so I really ought to give it a go. ;D
Caveman
|
|
|
|
Post by weverb on Nov 29, 2009 8:54:57 GMT -7
weverb, Glad you are keeping this thread alive I too would like to know the answers to your questions. With four drivers in a manifold, as opposed to two, there would be a greater degree of coupling. With regard to #3 though, I am certainly not the first to have an array and a manifold in the same room. A number of builders, including Chris have changed their arrays to manifolds. Did you perhaps mean, at the same time, in which case I do not know of anyone else silly enough to try that Can you expand on "coupling". Sorry Chris!  I forgot about your past travels. Since Jim's thread has been the "elder" documenting both, I forget to check your original thread. As for your last comment Jim, I think you are the perfect candidate for such a test. Not being silly, but having the space to set up both configurations. Consider it your punishment for not posting REW plots. ;D |
|
|
|
Post by chrisbee on Nov 29, 2009 9:06:51 GMT -7
Can you expand on "coupling". Coupling is where multiple drivers increase their combined output as a result of being placed close together and acting in unison. Now that's just plain cruel. ;D |
|
|
|
Post by weverb on Dec 10, 2009 7:19:59 GMT -7
I've thought about this a lot. ;D One should remember that manifold mounting places another driver cone opposite each driver cone. The opposing panel area is (usually) relatively small in comparison with the area of the opposed driver cones. The driver cones are not flat. (unlike a panel) Are at least partially acoustically transparent but are also usually moving linearly towards and away from the opposing cone. If only a little. Though the change in their spacing is effectively doubled. Which means that distance-related boundary effects between cones are constantly changing. Which means that any reflection effects between these opposing, irregular surfaces are very unlikely to peak at certain frequencies. i.e. No standing waves. Which reduces the likelihood of any boxy colouration. Further, the frequency of inter-boundary effects is likely to be well outside the pass band of the IB. If, like myself, you find you cannot sleep at night for worrying about these things then just make the opposing walls of the manifold non-parallel. Thereby removing all potential criticism of the manifold as an unwanted box in your IB system. In fact, provided that you have the woodworking skills, or a complete lack of them, there is really no need for any two surfaces of the manifold to be parallel. Just be sure to report back on your extended A/B/X comparisons with a standard square (box) manifold. So that we too may be "improved".  Don't think I'm making fun of you. I greatly admire those who seek improvements to the standard IB prescription. Most of the likely improvements (in my own limited experience) come from stiffening the box itself so that the driver baskets remain totally unmoving relative to each other and to their surroundings. Those seeking the best SQ could well look at concrete or a similarly stiff, massive and rigid medium (like bricks or slabs) for either a manifold or an array baffle wall. Those who prefer more normal, laminated boxes should consider stacking smaller, 2-opposed driver manifolds. Rather than aiming for one huge, but flexible, multi-driver, vertical manifold. (like mine) The thicker tops and bottoms of the smaller manifolds will stiffen the assembly better than adding shelves as an afterthought. It will also make the installation of an IB, containing a larger number of drivers, much easier. At least, compared with manhandling a single, huge and heavy box. I saw Chris post this comment in another thread and thought it might be good to add to our discussion. I wonder about your non-parallel surface comment and the hint Thomas dropped on us. It also made me come up with a new test suggestion for Chris. What if you were to redo your manifold two driver high by two driver deep? Or maybe actually split it into two manifolds (1 driver high by 2 drivers deep) collocated on the same wall? Just thinking out loud. |
|
|
|
Post by chrisbee on Dec 10, 2009 12:32:48 GMT -7
Hi weverb
Sorry, I can't really see the point of deepening the box from front to back as it will only add delay to the rear drivers.
I'm now considering an OB using the older, vinyl-coned Fs:32Hz drivers. This will return the enclosure volume/Total Vas ratio to normal for the newer IB drivers. This will reduce vibration in the baffle wall and double doors. The OB will consist of half of the present IB manifold with drivers opposing. Orientation will depend on REW and listening tests.
Placing the OB on the right will help to balance the offset IB on the left. Adding more bass output on the right should fill the present trough.
I have a deep mass of absorption behind the speakers so that can soak up some vibes from the OB. Each set of drivers will continue to be driven by one channel of the EP2500 with one channel of BFD EQ per set.
|
|
|
|
Post by ThomasW on Dec 10, 2009 13:38:17 GMT -7
I'm now considering an OB using the older, vinyl-coned Fs:32Hz drivers. Hint....... Put them in open baffles, one baffle mounted beside each main speaker. Sell trade-in or whatever your CX2310 for a CX3400. Find another stereo amp for these drivers run as 'woofers'. |
|
|
|
Post by chrisbee on Dec 21, 2009 3:24:42 GMT -7
Well I have had time to gather my thoughts further. I thought at first that it would be as easy as cutting the manifold in half. The free half would be placed over on the right as a 4 x 15" OB to balance the offset, 4 x 15" IB on the left. This would have several beneficial effects: There would be some thermal benefits from halving the cone area exposed to the cold air of the enclosure. The remaining 4 x 15" IB driver's total Vas multiplier would return to its former, more normal level. This would reduce acoustic and mechanical effects on the baffle wall and enclosure access doors. Using a standard 4 x 15" opposed driver manifold as an OB would provide a ready made, highly compact, folded, baffle-like structure with large cone area. One which also killed reaction forces. Its small footprint would just fit between the right hand speaker and my computer desk. Orientation would depend on experiment. I could "borrow" the SVS plate amp and run the 4 x 15" OB in series parallel, mono for 4 Ohms. I only need a few hundred Hz bandwidth for the OB so any doubts regarding the SQ of the SVS plate amp should not be a factor. The SVS plate amp would give me phase adjustment for the OB if it should prove desirable. An alternative to the manifold OB idea would be a plywood, OB "pelmet". This would hang from the junction between the ceiling and the 45 degree, sloping, back wall. It might appear a bit "top heavy" at first but would soon be ignored. It would be well away from foot traffic being situated right above the open stairwell between the stage and our seats. There is literally no room for open baffles lower down between the speakers and TV. The CX3400 might be allowed to slip under the radar once Christmas is over and diplomatic ties are re-established.  |
|
|
|
Post by saddlesore on Jan 1, 2010 11:46:41 GMT -7
I built two "v" shaped IB's (two Fi 18's in each) and couldn't be happier with the results.
Steve
|
|
|
|
Post by chrisbee on Jan 1, 2010 13:02:17 GMT -7
I built two "v" shaped IB's (two Fi 18's in each) and couldn't be happier with the results. Steve Hi Steve We'd be disappointed if you were. ;D Happy New Year Chris |
|
|
|
Post by weverb on Jan 2, 2010 13:01:12 GMT -7
I built two "v" shaped IB's (two Fi 18's in each) and couldn't be happier with the results. Steve Steve, Have any pictures available of the install/set-up? |
|
|
|
Post by jimcant on Mar 12, 2010 0:46:46 GMT -7
Hi Chris/Caveman/Marvin Some of the discussion in the Projects Section has started me thinking  The seed was your comment about decoupling of OB drivers, which Darrel Hawthorne does do with his builds. I don't know how effective the sponge rubber is but I guess it does help So I was thinking, what if one had an IB array and used a similar technique to decouple it from the walls? I currently have a 15" bass driver in my OB speaker baffle, and whilst it certainly generates some movement, it does not seem to detract from its performance.  I have built one stud wall so far and covered it with sheetrock, but still have to flush the joints. Being an inveterate fiddler, I was thinking of taking my drivers out again and mounting them facing in, instead of magnets in, just to see/hear what happened, presumably nothing. Then I got to thinking that I could use the manifolds as a test bed for a single driver array IB, by placing a panel over the manifold opening with a driver mounted in it. I could even use a panel that rested on the floor, and then use some sort of appropriate foam around the driver to seal the panel to the wall somewhat. The rear of the drivers would vent through the two open holes in each manifold, which is not ideal but would be OK for test purposes. I am currently thinking about what sort of material would work best for the "gasket". Just think, this might have been the solution for your array There might even be a way to harness that energy with some sort of piezo material  I probably ought to stop now Anyway, do you (or anyone else) think this area is worth investigating further, provided it has not been adequately covered previously of course. I have a WAF free zone so I am happy to be the Guinea Pig ;D Cheers, Jim. |
|
|
|
Post by ThomasW on Mar 12, 2010 9:49:59 GMT -7
|
|
|
|
Post by jimcant on Mar 12, 2010 19:08:29 GMT -7
Thomas W, Thanks for that link Much useful info, even if some is a bit over my head I had actually wondered about the merits of magnet mounting, so it was good to find that out too. Cheers, Jim. |
|
|
|
Post by twisterz on Mar 12, 2010 20:20:17 GMT -7
Cheers to all. This has been a wonderful read  . Now I'm going to watch Star Trek and enjoy my system. ;D |
|
