It began when we bought a house that needed renovating. The house also needed a big shed… so that’s what it got first. But after the shed was built, and time passed, I decided I’d better get around to doing the work I’d long promised.
That’s work like putting-up new plasterboard on the walls of the dining room and kitchen, replacing all the kitchen cupboards and bench-tops, painting, and laying ceramic floor tiles.
This, you must understand, is not work I find particularly interesting. It’s OK – especially when you’re learning new skills, but after only a short time it starts to lose its glamour. So, I thought while floor-tiling in the lounge room, what could I do to make this project a bit more interesting?
Hmm – how about built-in speakers? Like, subwoofers located in the crawl-space under the house, firing up through floor-mounted grilles? And then to complement them, speakers built into the walls?
In the past such ideas would have filled me with terror (what, tearing the floor and walls apart?) but these days, having become modestly proficient in flooring and plasterboarding, I reckoned it was all do-able.
Let’s start with the floor speakers.
My first thought was to mount a couple of drivers under the floor, with their enclosures being the whole of the atmosphere. That is, to use a genuine infinite baffle approach.
(If you don’t know anything about the ins and outs of speaker enclosure design, refer to the end of this story.)
If you look around the Web you’ll see some fantastic infinite baffle designs, where people have mounted drivers (often multiple, large drivers – like four 18 inch sub-woofers!) so that they work in under-floor or in-ceiling locations, pumping sound into the room through a short connecting duct.
The advantages of an infinite baffle design is that there is no speaker box to colour the sound, and with an infinite amount of air behind the cones, no stiffening of the driver’s suspension through the trapped volume of springy air.
The disadvantages are that without that trapped air, cone excursions can be large and so you need a lot of driver area (ie lots of big speakers) so that you can use less power and have smaller cone excursions - and yet still get adequate air movement.
In my application, I could think of two further disadvantages of the infinite baffle approach.
Firstly, mounted under the floor, the rear of the speaker is exposed to outside air – and so, during rain and fog, airborne moisture. With the crawl space well-ventilated, I could see the moisture degrading the speakers over time.
Secondly, and for me this one was the big negative, the subwoofer cones are not particularly good heat insulators (why should they be?) and so in winter, I could see my heating bill being even bigger than usual! After all, it’d be much the equivalent of having two big holes in the floor…
So how could these disadvantages – especially the heat loss one – be overcome? The answer is to use an enclosure – to box-in the rear of the speaker. After all, there’s plenty of room for a big box… or two.
But before going much further, I needed to bite the bullet and source the drivers that I was going to use in this underfloor location. I decided to use two drivers (ie a stereo pair… I am not much into home theatre and so the system is predicated around two channel use) and started looking hard for suitable speakers.
I settled on some car subwoofers – the JBL GT5-15. These are 15 inches in diameter and have a continuous power handling of 300 watts. I bought new items privately for just AUD$190 a pair.
With the speakers in-hand, and with full Thiele-Small parameters available, I could start computer modelling different enclosure designs. I’ve long used BassBox Lite – my copy is dated 2002 – and have had success designing subwoofer enclosures with the program. It’s one I trust.
So I fired-up BassBox and started modelling.
The first design I modelled was the infinite baffle approach. Using the BassBox ‘sealed box’ program option (but with an inserted volume of 999999999 litres!) the software indicated a -3dB point of 33Hz. Pretty good – in the real world, as opposed to marketing literature, 33Hz is very low in note.
Using the software, I then trialled a sealed box – picking an arbitrary volume of 300 litres. The modelled response changed very little over the infinite baffle - after all, 300 litres is a pretty big box! In fact, when rounded, the -3dB point stayed at 33Hz.
But what about a ported box? The design of this is more complex, but by using a volume of 315 litres and a box tuned frequency of 20.5Hz (achieved in the software with a rectangular port 200 x 100 x 360mm long), the -3dB point could be stretched waaaaay down to 17.6Hz! That’s another whole octave lower than with the other designs – at 17Hz you wouldn’t be hearing the bass, you’ll be feeling the house shake.
This graph shows the modelled response of the three designs. Red is the infinite baffle, yellow is the sealed box of 300 litres, and green is the ported box of 315 litres.
Note that these were never final designs – just indicative of what could be achieved with each approach. And the trial modelling showed that a ported box was best.
The next step was to work out how it was all going to be achieved. One of the biggest stumbling blocks was to source grilles suitable for floor use. I wanted them to be able to be trodden on, to be appropriately sized and to, as much as possible, not look weirdly out of place in a lounge room. They’d be inset in a neutral coloured, ceramic tiled floor. (The floor I was now tiling!)
I started by looking at ducted heating vents (they call them ‘registers’) but these were generally smaller than I liked. In fact, the biggest I could find with an acceptable style was only 14 x 6 inches – just half the area of the driver.
I then contacted grille manufacturer Rayson Industries in Melbourne (Australia) and found the company very helpful. It turned out they had some metal floor grilles that they were discontinued items – they were therefore cheaper than usual. These were available in a variety of sizes - I bought two 254 x 356mm grilles in a beige colour. These cost about AUD$70 each. The grilles are powder-coated steel, are very strong and look decent.
Next, how were the boxes to be got under the house? There is just a single entrance area to the crawlspace – a door about 550mm square. It would be possible - but very difficult - to build the enclosures under the house (there’s not even sitting-up room), so it would be best to make them outside and then insert them through the access door. That meant a maximum width/height of about 500mm.
So for a box volume of (say) 300 litres, the dimensions would be about 500 x 500 x 1200mm long. The overall speaker diameter was 390mm, and the required cut-out 365mm – so that worked OK with a face dimension of 500mm.
On paper that all seemed fine - but how would such an enclosure sit under the floor? What was the orientation and spacing of the floor joists? Observation and measurement showed that the spacing was too close (at 400mm) for the enclosure (and driver) to nestle between the joists. The box would therefore have to be positioned up against the underside of the joists, with an extension sealing the enclosure to the floor hole.
Hmm, now how to integrate the port? Typically, the port of a vented enclosure exits the baffle at a place other than where the main speaker is located. However, in my case, all the sound from each enclosure had to come through single 254 x 356mm floor grilles – there could not be a separate grille for the port. The vent would therefore need to work through the same floor opening as the driver.
So how to do that? I decided to have the port enter from the side, feeding air into the extension that connected the speaker to the floor grille.
Further design work resulted in this final configuration: 200 litres volume with two curved, right-angled 100mm diameter ports. Here’s a sneak preview of the top of one of the partially constructed boxes.
Next issue: building and testing the 15 inch underfloor speakers.