Line array cornu spiral?

[jaylgordon]

I've been reading about the cornu spiral BH design and am intrigued. I was wondering if there's any hope for a small line array with this design. In other words, instead of a single speaker right at the center, you are looking at a small line array (say, four 4.5-inch speakers aligned vertically) behind which is-- somehow-- a spiral horn system akin to the cornu spiral.

I've worked on the geometry of this arrangment, and the verticality of the line array does present some serious problems-- either the length is right but the contours are inconsistent, or the contours are right, but the lengths are inconsistent. That's because of trying to wrap multiple spirals around

o
o
o
o

instead of just

o

I am trying to do this with something akin to Tractrix dimensions, but admittedly this is more of an "art of fitting things together" problem than one which is generated by a scientific question.

Moreover, I assume this MIGHT be doomed simply because you can't get the advantages of the cornu spiral with four drivers, due to wavefront mismatches of all kinds that might arise. Still, I can't stop trying to sketch this thing. I did consider a 2x2 arrangement for the drivers, but for non-low frequencies, I assumed that would be a muddy mess in terms of wavefront consistency as compared 1x4 vertical line array.

Disclaimer: I appreciate the philosophy of NOT trying just to "use up" drivers one has lying around, but in this case, I can't stop thinking about how to, well, use up these eight 4.5 inch drivers I have lying around.

[HareBrained]

I think you need to look at the CH a little different. The CH design is actually symmetrical about a line. If you were to print the design, cut it thru the center, and rotate one half over the other, they would align. And since you're array is rectangular, the final shape is probably going to have to be rectangular.

You could also build 2 horns into each half or one horn into each quarter, just folding instead of spiraling.

[norman bates]

4 x 4.5" drivers in a flat vert array should not sound good..........

I have short array of four 5" framed full range drivers coming in a couple of weeks. But mine is convex, having a focal point 12' away.

That is another huge can of worms that you may not want.

The sweet spot will be 1 spot in the entire room (well depending on off axis response, maybe 1 couch wide). Don't slump either or your highs go away.

But you get more drivers, my 2 driver setup sounds better than 1 driver, clearer at normal volumes...............

And building such a beast is a pain, let alone the long pathway behind the drivers for the spiral horn.

But if you can live with the drawbacks and building it is not a hassle, go for it.

Norman

[jaylgordon]

Thanks for the replies. Yes, I do see that the Cornu is not a spiral per se-- which is why my drawings keep looking like variations on the swastika... As for the flatness of the array (vs. convex), convex makes more sense, yeah, but that would be a whole different set up. I like the idea of splitting the horns so there's one for each driver (or for two at a time).

[HareBrained]

Simply because it hasn't been posted here:

Seb's Cornu Spiral

I don't see why you couldn't simply stretch this design along a symmetric axis (10 o'clock - 4 o'clock is the axis I'd chose). It would probably require to be widened as well. But you could just build it and tweak from there.

There's also the option of rolling off the 4 drivers at ~2500Hz and adding a tweeter. Not a full range implementation but then full range and line array are generally considered mutually exclusive.

[jaylgordon]

Thanks for the tip-- I'd seen that one, and it does give a good idea of how to do the construction without tons of hardcore tools.

I'm starting to accept that the whole notion of a line-array cornu spiral horn is pretty silly. I did get kind of attached to thinking about how to work out the geometry, but at the expense of a lot of crucial factors.

Probably I'll work these each into their own experimental boxes and baffles, working from theory, just to learn more. I built the factory recommended Fostex 168E Sigma backhorns several years ago, kind of robotically working from plans without studying the theory much (except for the Tractrix curve, which isn't too tough to get).

Thanks again for the tips and responses.