John Kustin
04/12/2021, 5:57 PMsky130_fd_pr__res_xhigh_po_0p35 type resistors, so I've been using magic's device creator to find the appropriate lengths.
The computed resistance in magic is different than the resulting resistance in ngspice.
For w=0.35um and l=0.5um magic reports a value of 3.542kOhm, but ngspice reports a value of 1.950016kOhm (.op sim @ 27degC). I'm using @User's sky130_tests/test_res.sch as a testbench.
A couple questions I have about this:
Which value should I use?
Does magic use a reference temperature?
What are some possible (or known) causes for this discrepancy?
In sky130_fd_pr__res_xhigh_po_0p35.model.spice I see the subcircuit references `sky130_fd_pr__res_xhigh_po__base`; in sky130_fd_pr__res_xhigh_po__base.model.spice, why is the tnom parameter commented out?
I've included pictures to help show what I'm seeing.