This project started with a requirement for eight matching speakers with good phase characteristics for use in an Ambisonic surround setup.
In practice it would be cheaper and much less effort to just go to Richer sounds and buy four pairs of decent budget loudspeakers, but I was curious to see how this design could be made to turn out, and I very much wanted to try to use a single wide-range driver. I also wanted floor standers to save buying stands.
I've not gone into too much detail on the construction. As mentioned above, I'd recommend just buying some normal floor standing speakers instead.
I don't like passive crossovers. (So much for sound engineering judgements based on science and measurement!) Passive crossover units in speakers have always seemed to be a very poor way to go about the task of separating drive to the two or more drivers. Building electronic crossovers is a doddle but then you need all those extra pesky amplifiers - OK for stereo but not for an eight channel system. So I did lots of web searching for decent wide-range drivers that went high enough to cover the top end well and low enough so that the compromised low frequency response could be compensated for with a sub-woofer. And I found some. E.J. Jordan makes the JX92S which can be seen on thier website.
It's a shame that they cost so much though, and I could only justify the expense after some internal mental trickery. "Look how much you've saved by building the boxes yourself - Oh yes, you've convinced me."
Transmission Line Design
I once read a Babani Press book by Kapel where he espouses TX line design and wide range drivers. That's fair enough, but I'm deeply unconvinced about his enclosure design. I'm not an acoustics or a loudspeaker enclosure expert, but I suspect neither is he. So much so that I actually did a couple of slightly less than full-size trial speakers which were very close to his design to see if I could get a decent bass response from a couple of cheap-ish 6-inch drivers. I got the response I expected - which was somewhat worse that putting the speaker straight into a very small sealed chipboard box. I think he stuffs his line with far too much wadding. If you actually try this much wadding the tube is pretty much sealed. I also deeply dislike the way the speaker feeds into the box sideways and somehow he expects the air movement or sound to just couple nicely into the TX line tube sticking sideways off the end. I thought I would have a go at my own design using some of the principles he described though. A friend insisted that it's a great idea to taper the line as it goes along, so I elected to go that way too. I like the custom of putting a 45 degree ceramic reflector directly behind the speaker which directs the sound down into the line. This surely totally does away with the problems of reflections from the back of the cabinet coming back out through the speaker cone, as long as the area into which this wave is reflected isn't fully rectangular itself.
As you can see, my line spirals around and vents out of the side(s) via a number of holes drilled in the side panels.
I hate cutting up wood. But if you choose your material carefully you can minimise how much cutting has to be done. DIY shops sell 3/8" furniture board in various standard sizes, and I chose my dimensions so that I could use 6" wide furniture board for the front and rear panels, all the internals, and standard width board for the two left and right sides. This has a pleasant side-effect, that standard width 6" ceramic tiles can be used for the sound reflectors internally. The internal baffles are glued and screwed securely to the side panels. You need to use real chipboard screws for this and all the other screws. 3/8" furniture board is only just wide enough to take chipboard screws without cracking up, and normal woodscrews will simply break up the stuck together sawdust and become loose. When all the internal panels are in, the ceramic reflectors are hot-glue gunned into place. You need lots of glue and it helps the adhesion if you warm the tiles up first.
Half Built Speakers looking like so many Open University logos lined up in the garden - In fact this is clearly an artistic installation, I just didn't realise at the time, due to being crippled from leaning over the B&D Workmate guiding the jigsaw. I did six first off before doing the final two, and I can not remember why. I think maybe the local Homebase ran out of black furniture board. The drivers were then purchased two at a time, as funds allowed.
"Stuff enough in to stop it sounding boomy in the bass," is the usual advice. Well they don't sound very boomy even with nothing in. Is this because I've made a fully tapered line which disappears almost to nothing before the vent? It seems intuitive that this will be less of a pipe-like resonator than a line that is the same cross section all the way along. I've since removed the wadding as it seems to make no discernible difference to the sound. The vents are now 5 11 mm drilled holes in each side panel at the end of the line. These can be blocked up with blu-tack to check the effect.
It's so difficult to judge without actually getting them in an anechoic chamber and measuring it. Certainly I expected more and lower bass, even with these diminuitive drivers. When I check the electrical resonance it peaks at about 65Hz.
A picture of the finished article, speaker number 2 of 8 in the Ambisonic surround rig. Makes a great stand for quality lighting pieces too. Careful observers may note that the cropped furniture board ends are somewhat crudely finished with black gaffa tape. Oh well, "function before beauty" is a fair motto.
More Sensible Designs?
I think that you could rationalise this design a lot and have equally good or better performance by not going quite so overboard on the line tapering. I particularly think that the final taper section before the vent is over the top. A new design would probably have the very first section along the top of the enclosure the same area, and have the rest slightly smaller equal area. The first reflector would be compromised a little. These adjustments might make it more resonant and allow the bass to extend, which could be controlled easily with a bit of wadding in the line. The construction would be easier with no awkward angles, and there would be more same length bits of wood. Also, the internal butt joints would all be at 90 degress so they wouldn't require quite so much filling up with hot glue as mine did.
Here is a dedicated transmission line speaker website. There are lots of interesting designs, parts of which have clearly followed similar thought processes to mine. I'm sure there is plenty more transmission line speaker wisdom there.
Due to popular demand, here are the original log-book drawings showing measurements.
TX line speaker prototype insides from the side.
Small prototype TX line speaker side panel drilling plan.
Final TX line design insides, from the side.
Final TX line design side panel drilling plan.
Final TX line design front panel speaker cut-out.