Question on PSRR of an Op-amp

[sudhakarj.tech]

Hi all,
I am designing an differential to single ended opamp with high PSRR, since i seen the Allen hollberg book shows how to measure PSRR of an opamp there it is given one input of an opamp is connected to gnd and other is connected feedback to output and there is a supply ripple on it. Another i surfed from net is one input is connected with dc of vdd/2 and other is feedback to out and there is a supply ripple on it. I think all these two methods related to the opamp with NMOS differential inputs and what if it is the case for PMOS input differentials???

I have one more question is, since if i take the second case like one input is connected with vdd/2 dc voltage and other is connected feedback to output (weather the opamp is NMOS or PMOS choice is urs) but my circuit operates with 200mv to 500mv range of input voltage to opamp input. since i give >=vdd/2 and check the PSRR then i am getting good value but when i give in the range 200mv to 500mv i am getting almost 10db or towards zero db. Which one i need to follow and Which is the current operation since my operation range is 200mv to 500mv not vdd/2. so PSRR is ???? what to follow for general consideration????? what are the general techniques to improve PSRR?????

 

[checkmate]

One thing to note is the difference between AC and DC ground. AC ground may be vdd/2, and it only means that it is a rail with zero AC signal. So the 2 cases are essentially the same, and they apply to both NMOS and PMOS input stages.

PSRR is generally dependent on the input pair tail current maintaining a high ac impedance. You have to know how to derive the common mode input range. If the tail current goes into linear region, down goes its output impedance (and hence your PSRR). If you have trouble with low inputs, use PMOS input stages.Where you have sufficient headroom, use a cascode tail current.