@User: Full RC extraction is something of a hack and will continue to be until I have some time to re-write the extresist code so that it no longer depends on the .sim format, which is not sufficiently detailed to contain all the information that extresist needs to pull from it. Your list of commands is more or less correct; the only changes I would suggest are: (0) start with "extract do local" to keep magic from writing .ext files in the installed PDK area; (6) You do not need to do ext2spice here. (9) You do not need to set rthresh to 0; it is not referring to extresist-type parasitic resistances and is not useful in this context; (11) Do not set scale to "on", especially for the SkyWater process. The result I get from doing your steps above is an almost-correct netlist, except that due to the limitations of the .sim format, it cannot get any information about the substrate and well connections from the .sim file, and ends up with "error" nodes in the output. These "error" nodes refer to the substrate and would need to be cleaned up by hand, or by script. As for your layout itself: These standard cells do not have internal connections to substrate and well. So your layout has isolated and floating nwells, and no connection from ground to substrate. This undoubtedly makes the extracted netlist even worse. What you want to do is to abut the standard cells boundary-to-boundary as they are intended to do in a digital layout. That will connect the wells together. Then you should put a tap cell (tapvpwrvgnd) on either end to make sure that the well and substrate are connected.

Hi @User 1. I am trying to put in a nsubstratencontact tap cell but I keep getting this error N-tap overlap of N-tap contact<0.12um in one direction (licon.7).

@ibad1112: You can do it that way, but what I meant was that the standard cell library has a cell called tapvpwrvgnd (with prefix and suffix) that makes the correct contacts, and is designed for abutting to logic cells from the same library.