[sdiy] Simple discrete Unity-Gain Follower ?

[Don Tillman]

   > Date: Fri, 02 May 2003 11:45:53 -0500
   > From: Grant Richter <grichter at asapnet.net>
   > 
   > Don:
   > > There are a number of serious problems with feedback loops,
   > > even when they're working properly.  The most blatent is that
   > > the output is the amplifier gain times the error.  That's just
   > > philosophically wrong at a deep fundamental level.
   > 
   > This is inherent to the design of a responding
   > servo-mechanism. The "error amplifier" is a simple differential
   > input stage, and the action of negative feedback is to cause the
   > output to track the input signal as closely as the
   > servo-mechanism can. If this works perfectly, the output will be
   > an exact replica of the input signal. And for DC, it does work
   > pretty darn well.

But the amplifier has finite gain.  And for stability that gain falls
off at 6dB/oct.  So at higher frequencies there will be a significant
error voltage, you can calculate exactly how much, for a nonzero
output voltage.  And this is when the loop is working "perfectly".

The result is a very weird kind of signal coloration.  It's one that's
not found in nature, so it can be very striking.  

I agree that feedback is great for DC though.

  -- Don

-- 
Don Tillman
Palo Alto, California, USA
don at till.com
http://www.till.com