I have a circuit (oscillator), in which I would like to scale the main transistor (NMOS oscillator transistor), at the same time I scale the PMOS biasing transistor. The goal, is by the end, to have curves how the NMOS parameters vary with an increasing current (especially gm). This should be done for the three regions of inversion, in order to see the trade-offs.
- Wmin_p is the minimum W for PMOS Bias transistor and the same applies to the NMOS (Wmin_n)
- W_p is the actual value of the W for the PMOS.
As we sweep W in PMOS, if we divide W_p by Wmin_p, we get a constant "k". I would like to use this constant to modify the W of the NMOS, each time I do it on the PMOS.
I have tried ADE L with DC sweep, parametric analysis and I am not able to put expressions into the parametric analysis, so it can run an analysis based on a simple expression as Wmin_p/W_p.
Thank you all.
why do you not sweep "k" and calculate W_p and W_n from it?
In reply to Marc Heise:
Thanks for your reply.
If I sweep "k", how can I plot and save the data for each "k" (gm, Id, VDSsat...) ? "k" is the relationship of the W_pmos and W_pmos_min and it is needed to dimension the NMOS. Once I would like to have a constant density of current, so every time I increase the Width of the PMOS, I should do the same for the NMOS.
A piece of information is missing: PMOS is operating in strong inversion and NMOS in weak inversion.