Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site crystal.UUCP
Path: utzoo!linus!philabs!cmcl2!seismo!uwvax!crystal!pda
From: pda@crystal.UUCP
Newsgroups: net.analog
Subject: Re: Video A/D converters
Message-ID: <468@crystal.UUCP>
Date: Mon, 3-Jun-85 09:02:20 EDT
Article-I.D.: crystal.468
Posted: Mon Jun  3 09:02:20 1985
Date-Received: Tue, 4-Jun-85 00:55:05 EDT
References: <127@uvicctr.UUCP> <465@crystal.UUCP> <1473@amdcad.UUCP>
Distribution: net
Organization: U of Wisconsin CS Dept
Lines: 27

> In article <465@crystal.UUCP>, pda@crystal.UUCP writes:
> > 
> > A flash encoder is in effect a stack of 2**n comparators with a laser trimmed
> > resistor network for the reference voltage at each comparator.  The physical
> > layout of the board around these devices is critical for good performance. They
> > have a tremendous bandwidth and are integrating digital and analog electronics
> > on one board.  Watch out for noise to couple through the board grounds and many
> > other places.
> > 
> > 	Paul Anderson
> > 	University of Wisconsin
> > 	Computer Sciences Dept.
> > 	Madison, Wis 53706
> 
> Actually the board layout at 4 Mhz isn't all that critical. It
> isn't nearly the problem that noise elimination at 100Mhz presents.
> 
> Mike @ AMDCAD


Yes that is true if the part in question is a 4mhz part.  When you are using
a part that has a 100 mhz analog bandwidth you still have to be carefull with
the board layout to reduce ground noise coupling and especially for the stability
of the circuit itself.  If it oscillates at 100 mhz but you are interested in
the last digit of accuracy at 4 mhz what happens?

Paul A.