I am trying to match lab measurement of short circuit noise for my pipeline ADC.  In order to accomplish this, I set up a PSS/PNOISE/PAC analysis for the first 3 MDAC stages and the S/H stage of the pipeline.  I set PSS to have a beat F=fs/4.  This appears to work fairly well, but I have a couple questions.

First, the PAC predicts a 6dB voltage attenuation across the S/H.  Is this expected?  The S/H does have a return to VCM forced during the sample stage by shorting the S/H positive and negative outputs.

In the simulation I have 3 MDAC stages with a 2X gain per stage, resulting in a 8X gain from the output of the 3rd MDAC to the output of the S/H.  Referring the noise at the MDAC output to the S/H input is easy (divide by 8), but when referring to the S/H input (to match the lab data) should I include the 6dB attenuation of the S/H?

Thanks,
Scott

Yes, The output is very nearly 50% duty cycle.

Thanks, that's what I thought.  It also corresponds with the pnoise analysis showing 6dB higher when I use the time domain analysis as opposed to the sources analysis.

Scott