Jim Thompson a écrit :
On Mon, 12 Mar 2007 20:33:27 -0800, John Larkin
wrote:

On Mon, 12 Mar 2007 18:24:06 -0700, Jim Thompson
wrote:

On Mon, 12 Mar 2007 18:09:24 -0700, John Larkin
wrote:

On Mon, 12 Mar 2007 14:39:06 -0700, Jim Thompson
wrote:

In reply to....

From: "tiger66"
Newsgroups: sci.electronics.design
Subject: fully differential opamp implementation
Date: 10 Mar 2007 07:59:40 -0800
Organization: http://groups.google.com
Lines: 12
Message-ID: . com

See attachment.

...Jim Thompson
If the vcvs inputs were in antiparallel, and not grounded, it would
have common-mode rejection. That's just two independent, grounded
amplifiers.

John

Not the point. I was just jerry-rigging some Spice stuff to get it to
work for the OP.... with his idealistic FLOATING source.

...Jim Thompson

Will that even Spice with a floating source? With no DC path to
ground? The real-life equivalent is not knowing what the hell the cm
voltage is.

John



In the direct E-mail notes to the OP I made a number of suggestions
regarding using Rfudge ;-) to put in some DC paths to ground.

Spending more time thinking about it, the VCVC inputs should simply be
paralleled, and other circuit elements applied for DC paths. The
output tap simply establishes output common-mode.


With an added VCVS to avoid reflecting the output current to the CM
input :-)

--
Thanks,
Fred.