I heard CBT's for the first time last weekend - JBL's 2" driver straight column, in a small PA. What struck me more than anything else is that the design would be well-suited to surround channels. A long, narrow, thin column is fairly easy to hide. If the shading resistors aren't too complicated/expensive, a CBT surround design based on the Aura Whisper would be interesting. 16 of them will sum to more volume displacement than a Seas Excel W22, though shading would change that a bit. Still, it should be adequate in SPL and extension (both ends) for surround roles when up against a sidewall. A composite decking post with a wooden insert inside to rear-mount the drivers would make the cabinets no big deal, too. And easily painted to match the room.
Also, Rick Craig recently showed an interesting new design for a half-height CBT with a 45-deg arc.
Quite the opposite of the "triple Malcolm" solution I wrote about in another thread, in that these look low enough to fit a TV above. (And a bit heavy to mount on the wall at the ceiling/wall juncture.)
Still, I was really intrigued, until I learned he was charging a bit over $6k/speaker. That's a bit rich for my blood.
What are the downsides of a shorter CBT? Could one run, say, a quarter-circle arc from floor to front wall and get substantially all the benefits of a larger CBT, albeit with less output potential?
The main reason I was intrigued by CBT was that it solved many of a line array's problems. It provides uniform directivity above the control frequency, (in the direction of the line), eliminates side-lobes in the direction of the line, has a slim profile great for LCR placement flanking and above/below a TV or acoustically opaque screen, and because they make for a near perfect surround speaker if used as a grounp-plane CBT using a ceiling as the 'groundplane'.
They solve many of a straight line's problems, but come with their own:
1. If 2-way or more, the CBT will have crossover induced lobing at right angles to the main direction of the line.
2. Above the control frequency, sensitivity drops at 3dB/octave, or 10dB/decade, with some irregularities depending on individual driver size, spacing and driver directivity. That means if you cross to 1kHz to your tweeters, to keep the same SPL, you would need 20dB of boost at 20kHZ. That's 100x the power. Although spread through a large amount of tweets, it is non-trivial amounts of power. With curved arrays, the problem is having enough HF power, not LF power.
3. Driver shading takes away 3dB of headroom. Instantly.
4. Crossover complexity can be increased, and shading resistors must be purchased, unless many channels of amp power will be used.
The Aura whisper driver would require a lot of boost due to it's falling response up high if implemented in a line. My first project will use a 2" fullrange driver with a rising response as TV flankers, literally attached to the TV (flat screen TVs are not known for their good sound, the speakers do not even face the listeners). They will be 10 drivers per array, and I will shade each driver individually. Although the rising response of the drivers will help, there will be HF issues, as with my first curved array with 2" drivers:
The HF response drops off noticeably, and this is not a shaded array (couldn't take the 3dB drop in output). Not a huge deal for surrounds, but noticeable on occasion, esp w/ 5.1 music.
Due to the falling HF response problem, AVR power will likely not be enough for a CBT, but we'll see, as 2nd CBT project is an outdoor groundplane array with 5.25" drivers and arrayed piezo horns, to cover both a backyard and a deck above.
A shorter CBT has the downside of a higher freq control point, but the advantage of a higher freq control point as far as HF rolloff is concerned. I will post spreadsheets that calulcate freq control points, etc....and hopefully summarize everything with a 'how to put together a DIY CBT' type paper.
A floor to wall CBT is unecessary, as the drivers near the wall would receive little power. Not sure a curved array straddling 2 corners has been simulated, but it would be a great thing to do.....