Path: utzoo!mnetor!uunet!lll-winken!lll-lcc!ames!pasteur!zabriskie!spp
From: spp@zabriskie.uucp (Steve Pope)
Newsgroups: comp.lsi
Subject: Re: Magic extraction (scmos)
Message-ID: <967@pasteur.Berkeley.Edu>
Date: 24 Feb 88 00:59:58 GMT
References: <3446@super.upenn.edu>
Sender: news@pasteur.Berkeley.Edu
Reply-To: spp@zabriskie.UUCP (Steve Pope)
Organization: Postgres Research Group, UC Berkeley
Lines: 11

Having read Marc and Taso's explanation of their magic
extraction problem, it seems the root of the problem lies
in the fact that magic does not really treat the mosfet's
as four-terminal devices when it extracts.  The node
of the substrate terminal has to be inferred from side 
information if it is needed for verification.  In some cases, 
this side information is dependent on "correct" assignment 
of source and drain.

Any other analog magic designers out there?

steve