I've been thinking and reading about Line Arrays, wanting to build a pair. After reading the papers concerning the CBT concept and power sharing, I'm wondering if my idea will work...
It basically is a copy of sort...of the Peavey Versarray....only on a home theater scale. I have the ribbon tweeters to do the project...th Peavey Versarray tweeter being sold by Parts Express. I have not choosen a mid bass driver yet. If I do the proposed array the mid bass could become a 6 1/2 inch driver instead of the usual 4 to 5 1/4 inch driver. And pushing it, it mght even be able to be the Parts Express DA175-8. The base of each array could be a sub, if you waned to go that route.
o, what o you think? Crazy? not doable because of comb effects, etc? Just go the standard baffle route? I'd loveto beon the cover of PEs Quarterly Sales flyer....bt they won't do it with non-stock driver..Icould tell the it is the Fountek Neo CD 2.0
Remember, this is "back of the envelope" stage in this project...
Sorry for the blast, but it's a lot to say.
Not familiar with the Peavey array, but it's likely a "far field" array if it's pro sound.
Your proposed layout heads away from the ideal in two different directions.
Since they are ribbons, the tweeters will have broad horizontal dispersion, but very narrow vertical dispersion. I would be afraid that tipping them away from each other would lead to coverage gaps rather than comb filtering. Referring to the image on page 15 of Jim Griffin’s white paper, imagine the effect on the gaps seen at left if you were to tilt the tweeters away from each other. At some distance, these gaps will fill in, but the question is at what distance?
I don’t know of any software that would allow you to model this accurately, since all of the software that I can think of assumes parallel oriented drivers. IOW, a flat baffle. Note that this is distinct from vertically offset acoustic centers, which is accounted for in all (useful) modeling software.
Larger woofers will tend to cause more comb filtering. At low frequencies, a woofer radiates omni-directionally, but as frequency rises, the dispersion narrows. Think of focusing a mag light from flood to spot. This behavior is based primarily on driver diameter. In a line array, you want the woofers in the line to acoustically couple with one another. The closer they are to each other, the more they behave like one driver, lessening comb filtering. Pushed high enough, they will all comb- as dispersion narrows, thus larger woofers require a lower crossover point, since enter to center distance on larger woofers must be greater. Again, tilting them away from each other should exacerbate this effect. You can calculate this from the formulae given in Jim’s paper, but generally, 5†is considered to be a sweet spot between radiating area and diameter. Truncated frames help here.
This acoustic coupling is what causes the radiation gain that Jim mentioned in your other thread. The woofers behave more like one driver with higher voltage sensitivity. Since they are directional vertically, this does not happen to any great extent with ribbons. Acoustic coupling is positive interference between drivers while comb filtering is negative interference. The difference is in whether they sum acoustically in phase or out, with all points in between.
You will notice that the CBT’s have a very large number of small drivers for these reasons. Don Keel has a very good write up of these speakers at his site.
If you need more radiating area, use a longer line. This had the added benefit of making the line look more like a true line-source from an acoustics standpoint, decreasing floor and ceiling reflections. This has its own tradeoffs, but that’s another post. Power tapering helps, and CBT addresses it more directly.
Keep in mind that this is all theoretical based on what I have read over the years in Jim’s paper and on the boards. I have never designed and build a line array.
You may want to play around with “Tolvan Edge†(free) to see the effects of driver placement, directivity and baffle width on response. Edge will allow you to array any number and most shapes of drivers. The effects are going to be somewhat approximate owing to the different behavior of specific drivers (esp. ribbons) but will be close enough for your purposes at this point- likely very close. You want the radiating surface, not the frame size. Do the woofer line and tweeter line separately- it'll be a mess if you do them together.
As always, I am open to constructive correction.
Have you considered playing around with the Linus and your mindidsp? Could be instructive, and low cost $ and labor wise.
