lc vco switch cap

Hi.
I am running some simulations about LC VCO in spectreRF. In order to extend the tuning range without increasing VCO gain, a swithing cap architecture for frequency coarse tuning is used. (plz see the file attached for reference)

While runing the phase noise simulation, if the MOS of switch cap is off, the result phase noise will be dominated by the switch MOS. I dont know why this happen...
Isnt the switch MOS turned off?

Should I add some DC path to bias the internal node?


Thanks!

where is the schematic

khouly

I have met the similar case.
My idea is that the mos doesn't turn off throughly. The spectre will get a wrong convergence.You can have a try.
And I did so, and i get a resonable result after that.

Sorry...I forgot to attach the schematic....


To adelezy,

I had tied the node of switch control to GND for turning off the MOS.