Hello,

I have designed a two-stage Miller fully-differential amplifier, with cmfb.  It seems that all benchmarks that I need from opamp are met (BW, slew, gain).  I am having a problem with the common mode output voltage in a Gm-Opamp-C integrator.  I am using a -0.9V to 0.9V supply.  At dc, both the OTA and the opamp have 0.3V CM when connected together and separately.  The problem occurs as follows: when integrating the output CM slowly rises to around 0.75V for one end and 0.8V for the other end.  

When using the opamp in an Active-RC integrator the circuit behaves as expected with CM at 0.3 for both ends.  Also, a Gm-C integrator behaves as expected with correct CM.  Simulations are done in Cadence.

Any reasoning as to why this is happening would be very helpful.  If more information is needed, I would be happy to oblige.

thanks, Matt

Here is the waveform image.  If allowed to run longer, Vop and Von tend tend get closer to each other.

The blue wave form is the input, and the red waveform is the differential output (Vop-Von).  So the output should have CM of 0, like the input, assuming that the CM of output end is the same.