Hi.

I tried to simulate the pnoise for a 4Ghz cross-coupled LC CMOS VCO in SpectreRF. I have followed the instruction in the documentation (and change the maxsideband to > 7)

However, when I look at the noise summary, I find that it says the main contribution to noise is due to the MOS (rather than the LC tank) for different sizing of transistors I have used, incl the min size case that barely start the oscillation. I always have the feeling that the main noise contribution is due to the low-Q LC tank but SpectreRF is saying that it is mainly due to channel noise of other MOS. I just wonder has anyone faced this situation before? I have used the spiral model for L and large PMOS for varactors, so their noise should be accounted for as well....

Thank you.