Hi,
I have downloaded 65nm models for TT, FF,and SS from ptm website. I want to test these on a cmos inverter using hspice. Kindly tell me how should i proceed. It is for the first time that i will be doing such simulation so kindly also tell me what result should i expect.
Previously I have done Monte carlo of RC filter in LTspice, is corner analysis different from Monte carlo ? If yes, how?
I have downloaded 65nm models for TT, FF,and SS from ptm website. I want to test these on a cmos inverter using hspice. Kindly tell me how should i proceed.
Create an hspice control file which contains the netlist of the inverter, together with control statements for your analysis, e.g. a DC input sweep run. Use .alter commands to run the simulations with various corner models.
Sure it is: corner analysis runs with individual corner models, i.e. one single simulation run for each corner, whereas MC analysis uses a different setup with variation ranges for several individual parameters (like threshold voltages and their offsets). The MC mechanism selects random combinations of these parameters (within the given limits) for each run.
Both methods are for different purposes: the w.c. analysis uses a distinct parameter set for w.c. conditions regarding the process, the supply voltage, and the temperature (abbr. PVT), whereas MC selects and runs many random combinations of these PVT (and more) parameters.
So you can't say one is better than the other. Corner simulations should be run in any case, MC analysis whenever possible, especially for regenerative circuits.
I'd suggest to read your simulation tool's and PDK's docu about corner and MC analysis, otherwise search G00GLE & Co. for general explanation of these methods. Also, Analog Circuit Design textbooks will provide information about this subject.
I have all almost those books on analog (**broken link removed**).
I don't have access to any pdk (that's why using ptm models).
Thank you very much. Your replies have been really a great help.