Alex Liao
Member level 4
Questions regarding capacitance calculation of Cgs in Spectre Simulatior
I run simulation using Spectre with BSIM 4.5 model . Once the simulation finished I do a print DC to print all like Cds, Cgd and Cgs capacitance. I try to know how do those values come? Are they from model file or the equation inside the Spectre simulator?
My experiment of DC Print. I have attached my value here .
The following attached is the Spectre manual.
My question comes here:
1.Why the naming is so wired, only the last one, cdb, is for intrinsic and the rest are for total?
2.Why we have negative value for capacitance?
3.Why Cgb is not equal to Cbg? The capacitance between two nodes should be the same regardless of the direction of describing it.
I also did another experiment. I deliberately modify my layout and then extract and run simulation.
I made simulation on new extracted calibre view. I found the final cgs, cgd and cds did not change a bit.
Here questions arise:
4. Since Cgs is the total effect why it did not change in my DC print table due to my modification on layout? It should reflect the partial effects of what I have deliberately modified.
Using what kind of equation does the Simulator calculate the Cgs and present it in the DC Print Table?
It is highly appreciated that you can share your knowledge here even you are able to solve part of my questions.
Thanks,
I run simulation using Spectre with BSIM 4.5 model . Once the simulation finished I do a print DC to print all like Cds, Cgd and Cgs capacitance. I try to know how do those values come? Are they from model file or the equation inside the Spectre simulator?
My experiment of DC Print. I have attached my value here .
Code:
Post-layout simulation DC Print
Cgs Cgd Cds
-9.072f -2.943f 13.08aMy question comes here:
1.Why the naming is so wired, only the last one, cdb, is for intrinsic and the rest are for total?
2.Why we have negative value for capacitance?
3.Why Cgb is not equal to Cbg? The capacitance between two nodes should be the same regardless of the direction of describing it.
I also did another experiment. I deliberately modify my layout and then extract and run simulation.
I made simulation on new extracted calibre view. I found the final cgs, cgd and cds did not change a bit.
Here questions arise:
4. Since Cgs is the total effect why it did not change in my DC print table due to my modification on layout? It should reflect the partial effects of what I have deliberately modified.
Using what kind of equation does the Simulator calculate the Cgs and present it in the DC Print Table?
It is highly appreciated that you can share your knowledge here even you are able to solve part of my questions.
Thanks,
Last edited:
