NetNews Usenet Archive 1992 #31/NN_1992_31.iso/spool/rec/audio/17089

Text File | 1992-12-21 | 3.8 KB | 81 lines

Newsgroups: rec.audio
Path: sparky!uunet!spool.mu.edu!darwin.sura.net!mlb.semi.harris.com!batman!gmc
From: gmc@linear (Sir Gerald)
Subject: Re: Amplifier Slew rate
Date: Mon, 21 Dec 1992 18:33:06 GMT
Nntp-Posting-Host: batman.mlb.semi.harris.com
Reply-To: [e-mail]
Organization: Analog
Sender: [e-mail]
Message-ID: <[e-mail]>
Lines: 68

bellcore!gizmo.bellcore.com![e-mail] (Michael O'Dell) writes:

 >When speaking about slew-rate, it is important to
 >consider that the units usually listed is Volts/microsecond.
 >
 >The problem is that speakers are current-mode devices.
 >WAIT - you say, Ohm's Law says we can translate between
 >the two, and that a large VOLTAGE slew rate will obviously
 >imply a large CURRENT slew rate.  That would be true for
 >a constant, pure Ohmic load (signal into a perfect resistor),
 >but speakers are pretty-far removed from a pure Ohmic load.
 >In fact, most of them are reactive as hell - especially
 >various planar driver types, but good ol' EM cones can be
 >pretty weird as well.
 >
 >One must actually be concerned with the AC Current
 >transfer function if you want to really understand what your
 >amplifier is managing to assert into your speaker's drivers.
 >
 >While voltage slew rate is an interesting parameter,
 >and not without considerable merit as a rough rule-of-thumb,
 >it does NOT explain why one amplifier will drive some speakers
 >well and does poorly on others.

     A speaker is a passive linear device for which Ohm's Law
always holds, whether the driving source is a current or a voltage.
The fact that the terminal impedance is reactive rather than resistive
does not negate the linear relationship of voltage and current
expressed by Ohm's Law. This terminal reactance is not constant
over the frequency band, so for a constant source driving the
speaker, the source's delivered power to the speaker, and the
speaker response, vary over the frequency band. Perhaps this is
what you were thinking about? Whatever, Ohm's Law still holds,
and the terminal V/I characteristics are independent of the driving
source. The V or the I can be forced by choice of source, but the
other will be given by Ohm's Law.

     Getting back to slew rate, slew rate is only one of many parameters
the amplifier designer has to consider in order meet the output
performance specs. If the amplifier meets it's power, bandwidth,
and distortion specs, and is honestly speced, then the slew rate, as
well as the other design parameters, are sufficient. More slew rate
will not then make the amplifier sound better. An amplifier of 200 watts
rms from 20 to 20khz needs only 5V/us maximum slew rate to deliver
its full 200 watts into an 8 ohm speaker at worst-case frequency of 20khz,
and needs much less at all other frequencies. An amplifier
of 200 watts, 5 V/us slew rate, 0.1% distortion will sound better than
one of 200 watts, 100V/us slew rate, 0.8% distortion. It is silly to
take one of the designer's parameters, ignore all other designer's
parameters, and treat this parameter as a performance spec. You don't
hear slew rate, you hear distortion. If the design meets distortion,
power, and bandwidth goals, it is fine. I suspect some manufacturer
tried for a marketing advantage by trumpeting a non-spec to turn
it into a perceived superior performance spec.

 >Who knows? This might even explain some part of the observable
 >voodoo-cable effects.

     More correctly, the doo-doo cable effect. The people who hear these
"cable effects" only seem to correctly identify which cable is which
when the listener can also observe which cable he is listening to.
These same people have even heard the cable effects in the same cable
when they thought they were listening to different cables, when the
tester, unknown to the listeners, gave them an A/A test rather than an
A/B test.