Labyrinthine musings:

As usual I am just thinking aloud again:

Having failed with an OB sub I am still feeling an urge to do something constructive with the four, older, AEIB15 drivers. (Vinyl cones, big magnets, Fs 32Hz, Xmax~15mm, remember them?)

While the present IB performance is fine I still want to reduce the IB's Total Vas Ratio to a more normal figure. I would like still more impact on drums and bass guitar than I have at the moment. A more "Hifi" sound if you want to be really picky. My Mission 753F, main speakers have only four 4" diameter bass/mid drivers with only the bottom two reflex ported.

Which is where the four older 15" drivers could excel if given the opportunity. Ideally they should be moved from the big manifold to a central position between the main speakers. Where a higher crossover point could be used if necessary or found desirable.

I may have achieved a reasonable balance between the two sets of drivers with the BFD. But it is obvious that they are still very different in electro-mechanical character. The newer drivers are being handicapped in a much reduced enclosure volume to driver ratio. Which is rolling them off early. The older drivers can't do low bass without massive boost. So the older drivers have to come out of the tall manifold but still be used for something useful. I hated to have them just sitting in boxes while I listened to the new drivers. It was such a waste of (huge) potential. Which is why I built the big manifold in the first place.

Behind and above the TV and main speakers is a 45 degree, sloping wall. Above this wall is the roof itself. There is at least 4" wide continuous eaves ventilation from end to end of the roof.  With 12-16" of rockwool resting above the flat ceiling and sloping surfaces of my attic AV room.

I am now thinking of fitting the older sub drivers so that they exhaust backwards through the knee wall. Either from baffles or manifolds. This will depend whether I decide to have two drivers in front and two behind me in an identical arrangement. (this might need a separate amp and EQ for each pair) Or have all four drivers in front which is easiest.

The rear wave would find itself opening upwards into a very wide, and quite deep, wedge-shaped area above the insulation but still under the roof surface. Some sound will inevitably escape through the continuous roof vents but I doubt this will matter given our rural situation. There would be no air resistance (at all) other than that required to negotiate the upward bend just beyond the knee wall. So in some senses it is an IB but with an extended path to freedom. Hence the term "labyrinth" mentioned in the title.

I am thinking about thermal losses while still having my audio "cake and eating it". A very low position in the room will avoid the higher temperatures up near the ceiling due to stratification in our two story, open plan home. Open plan probably sounds a bit too posh for an open stairwell in the middle of my AV room.  This stairwell is what has been putting me off fitting an IB in the ceiling. I could not possibly tolerate the heat loss through the cones in winter. The drivers already sweat in the unheated IB enclosure if the magnets are placed outside the manifold.

Any shutter on the warm side of ceiling-mounted drivers would cause the drivers to cool rapidly to outside temperatures. With subsequent and massive condensation upon reopening the shutter! Seven feet lower down at first floor level should reduce this problem considerably. Moreover, the drivers will not be directly subjected to outside temperatures. Nor will they be radiating room temperature to the open roof above the insulation. A manifold could  provide an exit to the cold outdoors as small as 17" square per pair of drivers. Far better than having the driver cones fully exposed to the open roof space itself.

The baskets would be placed on the warm side of the wall so that they enjoy the heat input from the room's ambient temperature. Driver orientation makes no great difference to sound quality with subwoofers. Placing the baskets in the room allows a simple, flat, fine wire mesh screen to protect the driver cones from any birds or other wildlife which might get past the careful screening at the eaves. (I had a tiny wren exploring the IB enclosure one day!)

The list of advantages goes on: The drivers would find themselves placed in a dark, invisible, storage area behind the TV stand. No special finish would be required as an existing curtain drops in front of this area to conceal the few storage boxes hidden back there. The very low position at the throat of the large horn, formed by the floor and sloping ceiling, should enjoy considerable boundary gain. The central position should help to reduce the present troughs between the speakers and the present IB.

While no offset bass effects are noticed in the soundfield, with the present IB location,  placing drivers centrally can't do any harm. To avoid stereo channel confusion I think I will use the 300 watt Bash amp from my SVS cylinder with all its extra controls. Though it is limited to a 120Hz maximum crossover point with a 12dB/octave slope this should be enough to let the IB15s provide a bit of extra impact.  

Shrinking the big IB manifold down to four opposed drivers would let much more light in through the glazed doors. It would allow a stiffer manifold, closer to the floor, for yet more boundary gain and better SQ. The big amp could go back on top of the manifold out of the way. I  am also looking at ideas for a sand-filled manifold rather like St.Louis Bob's superb OBs. Or even cast concrete panels. Or paving slabs for shelves or manifold reinforcement, cladding or resistive mass. A four driver manifold allows much more space between the drivers for such fun and games without the manifold box increasing too much in size.

You might be asking why not move all eight drivers to this "very desirable situation" in the sloping wall? Well, the present IB is a proven SQ winner with just four drivers. The enclosure volume is ideal for four drivers but not eight. The peculiar quasi-IB/ labyrinthine/ bandpass arrangement, which I am discussing here, may not offer the same quantity or quality of bass.  It remains an unknown quantity until it is built and tested half to death. Eight drivers would also take up much more room and require four manifolds.

I already have enough ready-cut baffles to throw two compact, two-driver manifolds together to check if this idea actually works.

27June 2010: Update: I built a quick and dirty opposed 2 driver manifold out of 3/4" plywood but have yet to cut new holes in the construction of the "happy home". Instead I turned the box opening downwards to achieve a sort of seriously undersized, sealed box. This sounded louder and deeper than having it open to the room. I tried reversing the polarity relative to the remaining two in the big manifold but could hear no difference subjectively. Have yet to try REW on the new box. The two sets of older drivers sound much quieter than the four newer drivers in the big manifold when I mute each box in turn on the active crossover.

Having removed a small area of the house roof covering I was able to peer down into the insulation filled void from above. Just to remind myself what I did when I was building the new roof. Clearance between the inner side of the 4" external, block house wall and the present plywood knee wall is only about 7".  I can most easily add a new knee wall/ baffle further into the room to allow more clearance if necessary. I could use weld mesh to fix the insulation on either side to achieve a very wide and deep, insulation-lined slot to the underside of the roof covering. (and thence up over the top of the 12"-16" of rockwool roof insulation) Any leakage to the great outdoors (at the well ventilated eaves) would be on the opposite side of the house to the nearest neighbours. I deliberately opened a dormer window while playing a loud and bassy CD today and found sound leakage well down. Not the least bit problematic.

It is quite shocking how much room two 15" drivers with magnets outwards on a modest 12" wide box take up indoors. For greater compactness, I am sorely tempted to build a horizontal line array despite my previous negative experience with such long baffles. I am thinking that the driver assembly will be working against the bottom of the rafters, floor joists and the entire roof mass above. (rather than sitting in a flexible wall between flimsy, original studs in the former vertical array)  The bulky magnets and baskets can best be lost in the space under the 45 degree sloping ceiling behind the baffle. Though this does raise thermal issues with cold drivers in winter. I can easily add more mass in the form of paving slabs and local bracing if necessary. As well as fixing the baffle/ knee wall firmly to the surrounding timbers.

Avoiding the present, +16dB bass boost at 20Hz will reduce reaction forces dramatically. I confirmed this today when the temporary sealing boards started rattling over the empty driver cut-outs in the manifold. Removing the boost instantly silenced the vicious buzzing. I was then able to raise the boost to +6dB without further problems. An hour of listening to Bass Outlaw's CD "Illegal Bass" was both interesting and entertaining. I found it impossible to localise the bass to the new box even when I had my head right beside it. All of the bass seemed to be coming only from the nearest speaker two feet away on the other side of my head. A very odd sensation since I had already raised the crossover point to 120hz.

I built the second manifold and placed that too open side down on the carpet behind the other main speaker. It was time to fire up the new [2010] version of REW.

N.B. No effort was made to EQ the results. Only to roughly balance the various units and ensure they were all connected in the best polarity relative to each other. The manifold boxes were still much quieter than the remaining 4 x 15" IB but I adjusted the gain until they matched well enough for listening tests. The new boxes peaked at 90hz rolling away slowly to below 20hz. The various combinations appear in the image below:

The red trace is all subs together using a 80Hz 24dB/octave active crossover but no speakers.

Yellow is with the speakers connected but deliberately set out of phase for comparison.

Blue is the best match between speakers, 4 x15" IB and the two box subs. (2 x 2 x15")

The 180Hz and 280Hz troughs are reduced relative to the 8 driver IB alone though not absent.

There is more energy in the upper bass now. Which is exactly what was hoped for with music.

The question now is whether it is worth hacking holes in the sloping wall behind the speakers to allow the two new boxes to breath as quasi-IBs? Or should I build a new horizontal line array (IB) for all four older drivers? The two new manifolds owe me nothing in time or materials as they were thrown together from scrap 3/4" plywood and simply screwed together.

As the rear wall slopes at 45 degrees I have various options regarding the height of the final design. The higher I place them the greater the likely heat loss through the cones. Though this does bring them nearer to the plane of the main speaker baffles. If  I raise a line array just high enough it won't upset the present storage arrangements at all.

A simple (long) box with bottom and front at right angles would bring the driver baffle to the vertical. Triangular ends would seal the box to the sloping wall.  Internal triangular braces would help to stiffen the whole arrangement.

Placing the array box three feet off the stage floor would offer the greatest freedom for air movement between the sloping indoor wall and outer roof surface. I have a weird roof construction which I built myself to make room for lots of rockwool insulation. By overlaying a new set of rafters at a lower angle I gained plenty of space without needing hugely deep rafters. The original roof was 45 degrees and very probably thatched when built. The new roof surface is 40 degrees. The small difference in angle provides plenty of space for well ventilated insulation. The large overhang was designed to allow heavy external insulation but I haven't got round to doing this yet. This rough drawing is not remotely to scale: Green is 12-16" of rockwool.

 As can be seen; exhausting the box to the space above the insulation would constitute a true IB as near as matters. Not only is the free volume above the insulation very large in area but the eaves are also well ventilated. The only thing which stopped me going ahead years ago was worrying about heat loss though the driver cones. The 12" insulating blanket would have to be pierced to let the drivers breathe. The area where the manifold/array would sit would lie behind existing, concealing curtains. I could always have a hinged, super-insulated, drop-down flap to cover the driver cones when not in use, I suppose. Though I would greatly prefer the insulation on the cold side of the drivers.

Getting such an idea past the "Head Gardener" is the greatest hurdle to fruition. Her patience with my IB building is wearing rather thin. Nothing I have tried so far reduces her love for  the old SVS cylinder. At least this tension offers me some leverage.  I can always claim I am only building new and better IBs in her interest. Nah, I wouldn't buy that story either. :-)

Click on any image for an enlargement. Back click to return to the text.