Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 (Denver Mods 7/26/84) 6/24/83; site drutx.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!harvard!talcott!panda!genrad!decvax!bellcore!allegra!ulysses!mhuxr!mhuxt!houxm!ahuta!drutx!pmr
From: pmr@drutx.UUCP (Rastocny)
Newsgroups: net.audio
Subject: Re: Speaker Physics (dynamic tweeters)
Message-ID: <2447@drutx.UUCP>
Date: Tue, 9-Apr-85 15:19:39 EST
Article-I.D.: drutx.2447
Posted: Tue Apr  9 15:19:39 1985
Date-Received: Sun, 14-Apr-85 23:41:40 EST
References: <464@umd5.UUCP>
Distribution: net
Organization: AT&T Information Systems Laboratories, Denver
Lines: 44

First let me clarify why I don't like to throw a lot of math and formulas out
on the net.  I choose not to so that MORE people will understand what I'm
talking about, not just the EEs.

There is much to explain about how tweeters work than presented in the first
article.  There are two basic problems with dynamic tweeters: 1) the rising
voice coil inductance with rising frequency, and 2) the dispersion effect
described in the previous article.

How do you get around these problems?  You could use a small voice coil
(< 0.25") but that limits either the impedance (less wire in the VC) or the
power handling ability (smaller wire in the VC) of driver.  You could also
change the shape of the piston from flat to hemispherical.  This changes how
the piston pushes against the air and extends the piston's dispersion-bandwidth.
Most manufacturer's choose the latter.

Another technique used with leaf drivers is to increase dispersion in one plane
and reduce it in another.  By making the driver long and narrow to maintain a
"large" piston area, you can increase dispersion at higher frequencies in
the narrow plane.  This phenomenon is referred to as the slit effect.  The trick
is to make the piston move back and forth uniformly.  If the edge lags the
center, you get all kinds of phase distortion and a poor sounding driver.

The third problem with tweeters lies with first reflections.  Wavelengths in
the tweeter's operating bandwidth are quite short.  Anything projecting above
the piston plane near the edge of the piston can set up standing waves that
impair the drivers performance (screws, the edge of the driver assembly,
grills, other drivers, etc.).

         |<--d1--->|      |<---------d2---------->|
---|screw|---------|piston|-----------------------|grill|--

So in the above illustration, standing waves are generated whose wavelengths
are 2d1 and 2d2, 4d1 and 4d2, etc.

This is getting too long.  There's too much involved to explain in this article
and I must get back to work.

		Yours for higher fidelity,
		Phil Rastocny
		AT&T-ISL
		ihnp4!drutx!pmr