Biasing folded cascode opamp

[sherline123]

Hi,

When we biasing for folded cascode opamp, normally we will make the current MN3 = MP1 + MP7
But I saw some design actually make the sinking current of MN3 slight higher than current passed MP1 + MP7
Is there any reasons why to make so?

 

[frankrose]

https://www.edaboard.com/showthread.php?382670-biasing-currents-for-mirrors-in-folded-OpAmp
Similar discussion, slewing vs noise performance is the point.

 

[ashrafsazid]

If I understand correctly, if you see in the design I_mn3> (I_mp1+I_mp7), [so, I_mp3=2*I_mp1], that's only due to the leakage. "Ideally" I_mp1 = I_mp7 for a folded-cascode can be employed. But if you require higher slewing or faster settling for some applications (depending on the cap or output sampling), then simply employ more current at the output branch, This time it could be 2*I_mp1 = I_mp7 = I_mp3.