Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!panda!teddy!rdp
From: rdp@teddy.UUCP (Richard D. Pierce)
Newsgroups: rec.audio,sci.electronics
Subject: Re: Mercury Filled Speaker Wire
Message-ID: <4367@teddy.UUCP>
Date: Wed, 23-Sep-87 09:35:47 EDT
Article-I.D.: teddy.4367
Posted: Wed Sep 23 09:35:47 1987
Date-Received: Sat, 26-Sep-87 01:46:37 EDT
References: <3816@watdcsu.waterloo.edu> <578@uthub.toronto.edu>
Reply-To: rdp@teddy.UUCP (Richard D. Pierce)
Distribution: rec
Organization: GenRad, Inc., Concord, Mass.
Lines: 68
Xref: mnetor rec.audio:3387 sci.electronics:1401

In article <1953@tekig4.TEK.COM> brianr@tekig4.UUCP (Brian Rhodefer) writes:
>The analog guru at the next desk assures me that, with today's
>amplifiers, the dominant influence on the transient/frequency
>response of the amplifier/cable/speaker electrical-to-pressure
>transducing system is the series resistance of the speaker circuit.
>He mentioned this by way of pointing out how profitless it is to
>worry about small fractions of an ohm's worth of cable resistance
>in the face of (typically) 4 ohms of intrinsic speaker coil resistance.
>

In part, your analog guru friend is quite correct, in that the series
electrical resistance in the voice coil is the dominant electrical loss,
but, as it turns out, it is but one of many total losses in the entire
system, and is not the dominant influence on the transient/frequency
response of the entire system.

>
>[A description of what amounts to an active loudspeaker feedback system
> that includes the effects of cable interface problems....]
>
>Moreover, forces generated by currents passing through magnetic
>fields tend to be simple products of the currents and field strengths.
>A speaker's voice coil, then, should then have a linear current-to-force
>transfer function.  As I'd expect voice-coil force to translate directly
>into induced air pressure, it seems to me that a current-mode output
>signal would also be desirable due to a more nearly linear transduction
>of audio voltage signals into air pressure signals.
>

Unfortunately for your theory, voice coil force DOES NOT translate directly
into induced air pressure. At its very best, voice coil force (and, hence,
voice coil current) translates into volumn velocity. The effective sound
pressure level is a complex function of frequency, due to the great
variations in sounf pressure level with frequency. (For direct radiator
loudspeakers, the radiation impedance goes roughly as the square of the
frequency while the driver is operating in its piston region. Below
mechanical resonance, the speaker is unable to maintain the constant
volume velocity, which is why they roll off. ABove the piston band, the 
radiation impednace is a complex and only partially understood consequence
of driver size, shape, baffle loading, diffraction, etc., etc.)

There is also the issue that the force generated by the voice coil in
practical drivers is not a linear function of voice coil current, even
ignoring frequency-dependent effects. This is due to non-linearities in
the mechanical suspension of the driver, non-linearities in the magnetic
field, etc.

>Am I mistaken in thinking that this is exactly the kind of transducer
>that an amplifier/speaker system ought to be?
>
>There must be some reason, though, for amplifier outputs to be made
>the way they are;  could someone kindly explain why?
>

AMplifiers are designed as voltage sources, because the vast majority of
loudspeakers have a varying impedance, yet, over the limited range of
operation, produce constant sound pressure for constant applied voltage.
At low frequencies, at and around mechanical resonance, the effective
electrical impedance is much higher than the nominal impedance. (I am, as
we speak, looking at the impedance curve for a 10" woofer that has an
impedance of 83 ohms at resonance, compared to a DC resistance of 3.2 ohms,
and a nominal impedance of 4 ohms. Its output, given a constant applied
voltage, at resonance is but 3 dB lower than that in the midband, because 
its in an enclosure that gives it a Q of .707. Note that the current flowing 
through the voice coil at that point is 1/20th the current flowing through it 
in the midband, some 26 dB less, yet the "efficiency" is the same.)

Dick Pierce