From: "Jay Andersson"
Subject: Re: Zeroing differential amplifiers (AoE)....
Date: Mon, 23 Dec 2002 15:09:37 -0000
NNTP-Posting-Date: Mon, 23 Dec 2002 15:09:36 +0000 (UTC)
X-Newsreader: Microsoft Outlook Express 6.00.2600.0000
"Winfield Hill" wrote in message
> I'm not sure what differential amplifier configuration you're
> considering. The purpose of the circuit on page 393 is to zero
> out any dc input when the auto-zero logic signal is present.
> If your input stage (replacing U1) is a differential amplifier,
> the auto-zero opamp (U2) will zero any dc offset resulting from
> the input stage, without any regard to the nature of that stage.
I'm using the classic style as per fig. 4.18.
> Usually in differential amplifiers it's important to maintain a
> high CMRR at high frequencies.
I'm only running at a couple of kHz max.
> This requires both symmetrical
> circuitry and generally a high-frequency CMRR trim capability.
> Certainly one would not want to disturb these circuits with a
> extra input paths, but there is an easy way to add an auto-zero
> input signal, which is at the differential amplifier's output
> reference node. For example, consider figure 7.27, just return
> the two grounded resistor paths to the output of your autozero
> integrator amplifier, or to an attenuated, buffered version of
> the integrator output.
Do you mean 7.26? If so, I can see how that'd work. Assuming I want to keep
it simple, would a straight 250K from the auto-zero integrator to the
non-inverting input be OK? Yes, the CMRR would degrade, but would it be
> One other option for you to consider, zero-offset error chopper
> amplifiers, discussed on pages 415 to 420.
At the moment, I'd like to stick to something I might be able to get my head
round, if possible... ;-)