Lpc2000

It sounds like your inputs aren't truly differential.  You've 
referenced the V- input as a reference 90mV.... which makes it a 
unipolar signal, essentially.  A differential input won't have this 
type of reference as Karem mentioned below.   You could force the 
inputs into the 0-5V range by using a 2.5V ADC reference.  You can 
still you a differential input but will then want to make the input 
amplifier's common mode reference at 2.5V... or half of the reference 
voltage so that it's centered in the range.  
Edwin

--- In lpc2000@yahoogroups.com, "Leighton Rowe" <leightonsrowe@y...> 
wrote:
> The analog inputs and references look stable to me. No floating 
> pins. I've made the Vin- input closer to ground:
> 
> Analog input (always +ve)---> Vin+ 
> 0V                       ---> Vin-
> 3V                       ---> Vref
> 5V                       ---> AVDD/DVDD
> 
> ...but the same readings are still coming out. Maybe it's because 
> I'm using a "true differential" ADC. I heard the setup's different 
> compared to pseudo differential which I'm more accustomed to using.
> 
> --- In lpc2000@yahoogroups.com, "Kerem Or" <k.or@s...> wrote:
> > It appears there is a common mode voltage problem. Check if there 
> is a 
> > current return path between AIN-/AIN+ and AGND. What you observed 
> can happen 
> > if the signal source is floating with respect to AGND. As you 
have 
> noted the 
> > AD accepts -2.5V to 2.5V. But since the part is not bipolar, the 
> applied 
> > input signal must not be above VCC or below ground with respect 
to 
> AGND. 
> > i.e, -2.5V actually means 0V connected to AIN+ and 2.5V connected 
> to AIN-. 
> > It is not -2.5V with respect to AGND.
> > 
> > Just make sure voltage on any of the differential pins is in 
> between the 
> > part's common mode range.
> > 
> > Kerem