2022-11-13 19:11 Tags: # Juliens Control Crash Course Don't use just an inductor tee bias because any changes in the FETs will put it off. Use a feedback instead. Eyeball the phase margin - 180, 360, 0 degrees. How do we analyse this so it's more obvious? Use pz analysis command - select positive output node Gives you the exact position of all poles and zeros Filter out up to 30 GHz This gives you a pole zero plot.  Pole zero plot - resonance is due to complex conjugate poles. Miller Compensation: Inverting gain stage - pop a capacitance across it. 20pF is quite aggressive. Even down to a femto Farad. Will need to play around - parametric sweep it. Look at the pole zero plot as we sweep the compensation capacitor. Compromise - BW and Stability (as you increase the compensation capacitor) You need to do a pz plot for stability - no way that it is stable (it will definitely oscillate). Direct Plot > Main Form iprobe - in series with feedback. [^1] --- # References [^1]: