In message , Martin Schöön
writes
François Yves Le Gal writes:
On 17 Jan 2008 09:02:01 +0100, (Schöön Martin)
wrote:
Interesting. I get 37 hits on Bruno Putzeys on AES' web. Do you
remember which year or at what location (since that is usually
how one remembers at which symposium one did or heard what)?
Try http://www.ciaudio.com/ucd_aes.pdf
Thanks, I'll study it imminently.
I have also found quite a bit of literature on International
Rectifier's web.
The IR AMP5 design is quite nice, but it is a sigma-delta
self-oscillating design using the controlling feedback signal from
before the output filter. This gives very low distortion, but the Zout
of the reconstruction filter is in series with the loudspeaker and its
crossover network. I've measured errors in the amplifier output
response of over 4dB above 10kHz when feeding speakers whose crossover
networks present a very reactive load.
My professional interest is in the design of class-D amplifiers feeding
step-up transformers for 100V line PA systems. The load-invariant
property of the UCD technique copes admirably with the enormous
variation of load impedance that these amplifiers see. (Dependant on the
number of speakers connected to the amp output and their power tapping).
--
Chris Morriss