FPGA Central

Can anyone explain why is it that when using a sigma delta ADC with a
programmable gain larger than unity, we cannot connect the negative
input directly to ground. I read this on one of TI's application
notes (slaa283). They used a bias circuit instead.

-- A

On Mar 27, 6:34*pm, amerdsp <amer...@hotmail.com> wrote:
> Can anyone explain why is it that when using a sigma delta ADC with a
> programmable gain larger than unity, we cannot connect the negative
> input directly to ground. *I read this on one of TI's application
> notes (slaa283). *They used a bias circuit instead.
>
> -- A

My guess is that the PGA is a typical 4-resistor differential-to-
single-ended converter arrangement, with the internal "bottom"
resistor connected between the op-amp plus input and the internal CM
(common-mode). If this is true, then the common-mode voltage on the
input of the internal op-amp is a mix of the internal common-mode and
the voltage on the negative input pin, and as the gain is increased
the external voltage dominates over the internal voltage. If this
external voltage is ground then eventually the op-amp craps out
because the common-mode voltage becomes too low.

Another explanation is that people who live in Texas should stick to
what they know and stay away from analog circuits


Bob

On Fri, 28 Mar 2008 20:29:08 -0700 (PDT), Robert Adams
<[email protected]> wrote:


>
>Another explanation is that people who live in Texas should stick to
>what they know and stay away from analog circuits
>
>
>Bob

It's been an interesting week here in comp.dsp-land, and that puts a
pretty satisfactory end to my Friday.

Thanks, Bob, for that chuckle.

Eric Jacobsen
Minister of Algorithms
Abineau Communications
http://www.ericjacobsen.org

Robert Adams wrote:
> On Mar 27, 6:34 pm, amerdsp <amer...@hotmail.com> wrote:
>> Can anyone explain why is it that when using a sigma delta ADC with a
>> programmable gain larger than unity, we cannot connect the negative
>> input directly to ground. I read this on one of TI's application
>> notes (slaa283). They used a bias circuit instead.
>>
>> -- A
>
> My guess is that the PGA is a typical 4-resistor differential-to-
> single-ended converter arrangement, with the internal "bottom"
> resistor connected between the op-amp plus input and the internal CM
> (common-mode). If this is true, then the common-mode voltage on the
> input of the internal op-amp is a mix of the internal common-mode and
> the voltage on the negative input pin, and as the gain is increased
> the external voltage dominates over the internal voltage. If this
> external voltage is ground then eventually the op-amp craps out
> because the common-mode voltage becomes too low.
>
> Another explanation is that people who live in Texas should stick to
> what they know and stay away from analog circuits
>

Oh, c'mon. I've used their THS4021 in a design. That is a blazingly fast
opamp and very low in noise. So they must understand a thing or two
about analog. AFAIK that was done at TI and isn't a BB design.

--
Regards, Joerg

http://www.analogconsultants.com/