SNR in differential circuits

[edafisher]

Hi all,

I have a problem.

For a fully differential voltage amp , with the same input voltage magnitude as the single-ende version, we will have the same output magnitude, but twice the noise, thus SNR is 3dB lower than single-ended version.

For a fully differential transimpedance amp , with the same input current magnitude as the single-ende version, we will have the twice the output magnitude, but twice the noise, thus SNR is 3dB higher than single-ended version.

Am I correct?

Fisher

 

[edafisher]

any comment? or should i move this to "analog ic"? anybody...