Path: utzoo!attcan!uunet!cs.utexas.edu!uwm.edu!csd4.csd.uwm.edu!info-high-audio-request
From: ames!versatc!peter@rex.cs.tulane.edu (Peter Tapscott )
Newsgroups: rec.audio.high-end
Subject: Re: bi-amping
Message-ID: <9100@uwm.edu>
Date: 23 Jan 91 13:56:04 GMT
Sender: news@uwm.edu
Lines: 125
Approved: tjk@csd4.csd.uwm.edu
Originator: tjk@csd4.csd.uwm.edu

[Many lines of reference deleted. -tjk]

I was trying to not write about this, since it seems like an interseting
topic that could require a lot of time to discuss thoroughly.  Oh, well,
here goes:

I see you having three options:
(1) Simple Bi-amp with two stereo amps (or 4 mono amps)
(2) Low-level crossover, probably an active crossover
(3) A hybrid of 2 & 3, probably a passive low-level crossover.

To explain:
(1) This means dedicating a stereo amp to each speaker.  The crossovers
in the speaker are used to isolate the drivers from the parts of the
frequency spectrum that are not needed, as they do now.

The advantage here is somewhat similar to bi-wiring if you only consider
the output impedance of the amp.  What I see being of more importance
is the greatly reduced current requirements of the unused frequency
portions.  The amp still has to produce the VOLTAGE it did before
bi-amping (as Norm points out above), but the current requirements
are virtually zero for frequencies outside the band allowed by the
speakers crossover.

That is, the bass amp only produces current for 300hz & below, the
mid-tweeter amp only produces current for 300hz and up.  Each amp 
still has to produce volatge for the entire spectrum.

Back when I used to make a lot of prototype amps, it was really
easy to make an amp that measured perfectly (as read by my crude
instruments) with no current requirements.  Add a real load,
though, and all sorts of distortion showed up.  This was particularly
true for crossover distortion.  

(2) With an active crossover, you can tweak a lot of parameters,
such as crossover frequency, slope, (perhaps) Q, type (Butterworth,
etc.)  The internal crossovers in the speakers are disconnected.
Your power requirements will decrease with frequency, since
the amps now have to only produce the voltage and current for
their specific frequency ranges.  A modest sized tube amp for
the tweeter section might sound great.

(3) This is a hybrid I thought of from your description.  This
method still uses the crossovers in the speakers.  Assuming that
your preamp has a substantially lower output impedance than the 
power amp input, you could use a simple passive filter of one
or two poles, using high quality capacitors and resistors.

I would place the frequencies of these filters so that the
crossovers in the speakers completely control the drive to
the drivers, but so the amps don't have to produce voltage 
for the entire audio spectrum. 

For example, the bass amp could have a frequency response
of 500hz, so the 300hz crossover in the speaker is providing
the rolloff, not the crossover I describe.

ADVANTAGES AND DISADVANTAGES:
(1) AD: Simple Biamp: any one can make it.  You need a good Y
connector if your preamp has only one output, but this could
be made easily.  (You can buy low-fi Ys at Radio Shack.)

This method is recommended by Vandersteen for the 2Ci, BTW.

The NAD integrated amp manual I saw that suggests upgrading by
Bridge-amping the internal power amp to one speaker, and 
buying a second power amp to bridge to the second speaker,
I would be interested if better sound doesn't result from 
simple bi-amping as described here.  (Perhaps a reader has
this setup and could run the experiment?)

DIS: Requires four identical amps.  The mid-tweeter amp
will just be idling most of the time, in terms of power 
output.

(2) AD: Active Crossover: Really a nice way to match drivers 
that have different characteristics.  you don't have to
use identical amps: amps can be tailored to the drivers.

DIS: The B&W 801s have very carefully designed crossovers.
They might even have some form of passive frequency
compenstation built in (I don't know).  These crossovers
have been proven with instrumentation far beyond what
most audio experimenters have, so I would use caution
in trying to redevelop the crossovers.

Also, a high quality avtive crossover is expensive.  This
is an expense you could put into buying better amps, or
even records & CDs.

(3) AD: Hybrid technique: as with the active crossover
method, identical amps are not required.  If you want a 
transistor behemoth on the bottom end and a tube gem for
the high end, fine.  With a low crossover, like 300hz,
though, I would not make the mid-tweeter amp underpowered.
This contains most of the information you listen to.
(For example, telephones have about 300-3Khz response, and
we can clearly understand the spoken word even though
the fundamental frequencies for many voices are missing.)

You must depend on a well designed crossover, but that 
is not a problem with the B&W 801s.  If you were talking 
about <$500, it may come into play, but not for the multi 
kilo-buck speakers.

DIS: you have to make the crossover.  This is easy enough
for me: a couple resistors and capacitors, but if you
aren't an electrical engineer, and a pretty competant
technician, this could be a problem.  Of course, if you
need help, I bet you could be able to hire some consulting
help from the net (not me).  

BTW, I use a slightly different version of #3 for my
subwoofer system.  The crossover took about 20 minutes
to design and about 20 minutes to build.  Scrouging
parts was probably another 20 minutes.  But I never said
that it is a work of art.  (2-pole passive filter.)

Will the original poster please post the results of his
experiments?  Thanks!
-- 
|----------------------------------------------------------------------|
| peter@versatc.VERSATEC.COM     -OR-    {ames|apple|sun}!versatc!peter|
| Peter Tapscott - Xerox Engineering Systems, Versatec Products        |
| 2805 Bowers Avenue, Santa Clara, Calif                 (408)982-4235 |