Hey everyone, so I would like to dive into the transistor mismatch analysis for this PDK. From this classical paper on transistor mismatch (https://ieeexplore.ieee.org/document/572629), there are two major contributors to mismatch, one is due to

Vt

and another is due to the charge mobility (electrons

\mu_n

and holes

\mu_p

), and their mismatch equation is attached as the picture. In short, where could I find that parameter

Ap

for

Vt

and

\mu

? I know there is this slope file here: https://github.com/google/skywater-pdk-libs-sky130_fd_pr/blob/f62031a1be9aefe902d6[…]/cells/nfet_01v8/sky130_fd_pr__nfet_01v8__mismatch.corner.spice But I am not 100% sure if they are really what I am looking for. For example, for the

Vt

there are two different slope values; and I do not see any ones related to the charge mobility. Any idea on this? Thanks.

Compare it with the TT corners files https://raw.githubusercontent.com/google/skywater-pdk-libs-sky130_fd_pr/f62031a1be9aefe90[…]36/cells/nfet_01v8/sky130_fd_pr__nfet_01v8__tt.pm3.spice

* SKY130 Spice File.
.param sky130_fd_pr__nfet_01v8__toxe_slope_spectre = 0.0
.param sky130_fd_pr__nfet_01v8__vth0_slope_spectre = 0.0
.param sky130_fd_pr__nfet_01v8__voff_slope_spectre = 0.0

+ toxe = {4.148e-09+sky130_fd_pr__nfet_01v8__toxe_slope_spectre*(4.148e-09*1.0*(sky130_fd_pr__nfet_01v8__toxe_slope/sqrt(l*w*mult)))}

vth0 = {0.5190093+sky130_fd_pr__nfet_01v8__vth0_slope_spectre*(sky130_fd_pr__nfet_01v8__vth0_slope/sqrt(l*w*mult))}

+ voff = {-0.1052686+sky130_fd_pr__nfet_01v8__voff_slope_spectre*(sky130_fd_pr__nfet_01v8__voff_slope/sqrt(l*w*mult))}

Thanks, I’ll take a closer look to that. However, did they just simply lump the charge mobility mismatch and gate oxide mismatch together and put it as the gate oxide mismatch (which is fine as muCox always appear as the product), or they didn’t actually do any charge mobility mismatch measurements at all?

I don't have the slightest idea. I think no charge mobility has any variability in the model files. Not for monte carlo.