Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!seismo!lll-crg!lll-lcc!vecpyr!amd!amdcad!decwrl!sun!saber!imagen!kchen From: kchen@imagen.UUCP (Kok Chen) Newsgroups: net.audio,net.analog Subject: Re: Amp designs (Re: current-dumping amplifiers) Message-ID: <481@imagen.UUCP> Date: Tue, 26-Aug-86 12:19:46 EDT Article-I.D.: imagen.481 Posted: Tue Aug 26 12:19:46 1986 Date-Received: Wed, 27-Aug-86 21:36:28 EDT References: <5975@alice.uUCp> Organization: Imagen Corp., Santa Clara, CA Lines: 30 Xref: mnetor net.audio:2419 net.analog:843 >> ark@alice writes: > > My understanding is that "current dumping" is a trademark of the > ... amplifiers (QUAD 403). > Right on! This is the first proper description of "current dumping (TM)" that has appeared so far. In fact, what has described is the way Quad does FEED-FORWARD. The error signal from the cheapo high-current section is fed to a lower power section to cancel at the speakers (instead of feeding the error signal BACK to the input of the high-current amplifier). I, for one, would be interested in hearing more on feed-forward power amp topologies. One idea that I kicked around has been to difference the main and feed- forward paths by feeding the output of the main driver to one speaker wire, and the output of the error-signal amplifier to the second speaker wire (like a diff amp). Two possible flaws in this: 1) won't work with electrostatics that require single ended inputs, and 2) the finiteness of the first driver's output impedance will introduce some positive feedback to the second driver. Another idea: the feed-forward signal (the input to the second amp) can come from an auxiliary cable that goes straight to the speaker terminals (like in remote-sensing power supplies), instead of to the output of the first driver. I haven't tried out any of these ideas yet, except on paper (I do have a spare DH-120 that I don't mind modifying and blowing up, though, if I ever find time :-). Anyone see major flaws in the logic? Kok Chen Imagen Corporation (we don't manufacture amplifiers; in fact, we don't even do anything remotely analog!)