2stage fully differential opamp output distortion

[tompham]

Hi all

Im desinging 2 stage fully differential opamp with continous cmfb. My design using smic18mmrf with VDD=3.3v and input cm = output cm = 1.65v. I test transient with apply sin wave 20uV peak to peak
to see how output behave. From the waveform (see attached) , the output distortion at below output common mode. I have hard time to figure the problem even the opamp has good output swing. Anyone can help me with this. Thanks a lot

 

[FvM]

You see in picture 4 what the actual output swing is (either if it's considered good or not). Shouldn't the CM bias be centered to the output voltage range?

The other question is why the output clips below 1.5 V. I fear , this can't be seen from your selected circuit snippets.

 

[tompham]

Hi FvM

Thanks for your response. I use common mode reference = 1.65v = vdd/2
This is clip below 1.5v from my simulation

 

[Dominik Przyborowski]

It's look like your current sink controlled by CMFB has very poor performance. Add some degeneration to them or cascodes and check what will happen.

 

[BradtheRad]

The clipping at 1.5V reminds me of typical op amp behavior.
Example, LM741, whose output cannot go closer than a couple volts above its bottom supply rail.

 

[Dominik Przyborowski]

The clipping at 1.5V reminds me of typical op amp behavior.
Example, LM741, whose output cannot go closer than a couple volts above its bottom supply rail.

Yeah, but not for cmos process with 3.3V supply ;-). I gues that diff pair is loaded by simple current source maded on one nmos so the lower rail limitation is equal to VDSsat voltage around 0.2-0.3V.