Negative resistance of LC VCO decreases with width...?

[haadi20]

Dear all,

I remove the tank from a cross-coupled topology and see the simulated real part of the impedance seen from the cross-coupled pair. As expected, it has a negative value.

Then i run a sweep by changing the # of fingers of the transistor but the results are opposite to what is expected. The negative resistance decreases with increasing width of the transistors?

Can someone please explain the reason? See the graphs for swept simulation of number of fingers.

 

[Milad-D]

The negative resistance of the Cross_Coupled topology is -2/gm and gm is sqrt(Iss*Un*Cox*W/L) in which Iss is the bias current. Therefore as you increase the W the gm will increase and this leads to lower negative resistance.