How can one estimate an acceptable Vov and Vds for a transistor such that Vds >= Vov is satisfied?

[kimmyo6]

Hi!

I always have this doubt when the circuit gets bigger and I end up going to the simulator almost blindly because I don't know where to start. Say you're designing an differential opamp with active pmos load and a nmos for the tail current. Assuming you want all 5 transistors in saturation, how do you begin your calculations? How is it done for bigger and more complex circuits? Allen makes a step-by-step hand calculations for a miller two stage op amp in his book and it seems to work fairly ok, but what about all other circuits? A latched comparator? A bandgap? Literature seems so confusing because everyone has a different way of doing it, unfortunatelly. If someone could shed a light, I'd appreciate.

 

[dick_freebird]

First, lose the notion that every device must be in saturation. It's untrue and impossible for any complex amplifier with wide swinging voltages.

Determine which ones matter most, first, and decide also what's good enough for the combined dimensions of performance because every band-aid is also baggage and at some point "more" stops being "better" and even earlier, "worth it".