Path: utzoo!attcan!uunet!mailrus!cs.utexas.edu!uwm.edu!ndcvb.cc.nd.edu
From: przemek@ndcvb.cc.nd.edu (Przemek Klosowski)
Newsgroups: rec.audio.high-end
Subject: Re: (very) Unconventional Speakers
Message-ID: <2712@uwm.edu>
Date: 2 Mar 90 17:11:19 GMT
Sender: news@uwm.edu
Lines: 41
Approved: tjk@csd4.csd.uwm.edu


In article <2700@uwm.edu> ti@altos86.Altos.COM (Ti Kan) writes:
>
>In  <2678@uwm.edu> aipp@castle.edinburgh.ac.uk (Pavlos Papageorgiou) writes:
>>	Has anybody out there investigated, or at least considered, the
>>concept of a 'digital speaker'?  --- meaning a method for driving the
>
>I guess the question is *why*.  I don't think such a device will
>provide any real performance advantage over the current amplifier/speaker
>interface.  Going digital just for digital's sake is not what I call
>sensible engineering!  Modern power amps typically have vanishingly
>low noise and distortion, and has extremely wide bandwidth.  

Correct!

>All the
>problems that exist today in this setup are mostly in the speakers
>(i.e. phase shift from the drivers and crossover network, distortion
>from modulation and cone-flexing, power handling limitations, dispersion
>anomalies, resonances, etc, etc.  None of which can be fixed by driving
>the speaker digitally.

But that is the whole point---maybe by going digital all the way to the
drivers you can avoid crossover networks and phase problems coming from
non-resistive (i.e. having imaginary part) impedance of the whole system
(wires, xover, coils and cones). People tried to do closed loop servos
connected to cones, but it didn't catch (maybe because these were velocity
servos, and needed integration to get back displacement, all was done in 
analog and wasn't precise? Opinion on those?). Doing closed loop in digital
may be the thing to do... a speaker with Motorola 56xxx DSP inside...
Remember that the speakers are the WEAKEST by far link of a HiFi chain
(no wonder they are called SPEAKers :^).

>Ti Kan                                                                  \\\
>vorsprung durch technik!                                                 \\\
>Internet: ti@altos.com                                                /// \\\
>UUCP: ...!{sun|sco|pyramid|amdahl|uunet}!altos!ti                    ////////\

			przemek klosowski (przemek@ndcva.cc.nd.edu)
			Physics Dept
			University of Notre Dame IN 46556