[sdiy] Dynamics and speakers, was:Advice

[Thomas Dunker]

On Wed, 4 Feb 2004, Dave Krooshof wrote:

> Ideally it's that same with speakers as with microphones.
> You want them to be point sources. Any surface bigger
> then the wavelength is hard to control.
> There's a sound to your speaker just because it has a size.

> To make a tiny omni directrional speaker with hardly any
> mass is "easy". Make a tesla-coil, tune it to 44100Hz or
> double. Then do AM on that wave with the output of
> your cdplayer/computer/synth. Bigger or smaller sparks
> will produce airwaves. This is allready done, and it doesn't
> sound that bad at all. Beautiful transients!
> Pity about the ozone though.

 Actually this idea of the ideal speaker being a point source has a big
problem - I don't mean to discourage anyone into plasma speakers, but
talking about theoretical "ideals" when it comes to speakers is kind of
treacherous ground. The point source may be an *acoustical* ideal, but in
practical terms the problems are painfully apparent, as Reto Luigi
Andreoli pointed out in an article in HiFi Scene Schweiz some years ago:

"The "ideal", point source loudspeaker will have zero efficiency and
100% distortion."

 Actually, as good as plasma tweeters may sound, they have, quite
expectedly, a dense spectrum of intermodulation distortion (a lot of it
in the ultrasonic range, no doubt). With next to
no surface area (like most modern speakers...) they must try to make up
for it with increased "amplitude" (think of it as a very small pulsating
sphere) and the formula for frequency modulation distortion dictates that
this distortion increases as the product of bandwidth times peak velocity,
 As the area of the vibrating piston approaches zero, so must the velocity
approach infinity if we are to have any sound at all...hence the "100%"
distortion.

 I get your point, though. Given the ever shrinking size of speakers
today it's clear that they try to approach the point source ideal.
Unfortunately, with increasing distortion and lowered efficiency as a
result. This problem was worked out in theory by acousticians decades
before the first practical speakers were even invented, so it's not
exactly news anymore.

 Also, centuries of development of musical instruments provided a lot of
empirical knowledge of acoustics and so for all sound sources EXCEPT
ANYTHING INVOLVING LOUDSPEAKERS it's considered OBVIOUS that the
physical proportions of the sound source or instrument relate directly to
the wavelengths involved and/or the volume of sound produced. ONLY
speakers, expected to be able to emulate ALL conceivable natural or musical
(or synthetic) sounds, are also supposed to be small enough to fit into a large
shopping bag! My argument is that this is a result of logistic/economic
priorities rather than sound engineering in accordance with known theory. It's
quite easily proven.

Thomas Dunker