"TonyP" wrote in message
u
"Don Hills" wrote in message
...
In article ,
(Curious) wrote:
Probably having many dynamic super-tweeters would do the job. If
their are 140 speakers each playing at 1 dB, you get around 140 dB
if they are facing the same direction. Right? Wrong?
Wrong.
1 tweeter = 1 dB.
2 tweeters = 4 dB.
4 tweeters = 7 dB.
8 tweeters = 10 dB.
16 tweeters = 13 dB.
32 tweeters = 16 dB.
64 tweeters = 19 dB.
128 tweeters = 22 dB.
Close enough to 140 tweeters for you?
The 3dB increase each time is debatable, it depends on whether you
get phase coherence or not, but it's close enough for this example.
I think it's safe to assume 128 real tweeters will not produce phase
coherence due to the physical distances involved.
This problem could be addressed by adjusting the phase/frequency
characteristics of the drive to each speaker. In these days of cheap-but
powerful DSP and power amp chips, it's not totally unthinkable.
However, your idealized example correctly points out that the
price-performance of very large arrays sucks. Yet, when you've mounted the
highest efficiency drivers you can find on the most suitable waveguides you
can find or make, multiple-speaker arrays are all that is left.
There are also amplitude-only alignments like the N=5 Bessel array that can
produce an array that essentially behaves like just one driver, but with
greater power handling capacity.
Effective Performance of Bessel Arrays
Author(s): Keele, Jr., D. B.
Publication: Volume 38 Number 10 pp. 723·748; October 1990
Publication: Preprint 2846; Convention 87; October 1989
I recently built a N=5 Bessel array out of small long-stroke, "full-range"
drivers and it pretty much worked as claimed.