Hi ,
   I am simulating a fully diff Switched Cap Integrator. The input sample caps are 1p each. The amplifier has Adc=50dB,f3db=160KHz and the input reffered PSD of 1f /Hz. My clock frequency is 2MHz.

    I did a PSS + PNOISE to get the input reffered totoal noise. My PSS
beat freq=2M
N harmmonics=100
Facmax= 250M

My PNOISE has  sweep range of 1 to 2M. Reference side band is 0 ofcourse. Noise type is "sources" .

When I run the sim my total input referred noise power (intgrated from 1 to 1MHz) is 90n (much higher than I expect).

I also have a noise less model of my op amp and when that is used in the sim instead of the real amp then the input referred noise power is just what you expect from 1pF sampling caps ( ~16n).

Could anyone please give me some suggestion?
thanks much
shaf