@Stefan Schippers I am trying to plot gm graphs for gm/Id simulation but I cannot see plots when I type the plot command in simulation block. I can see the plot when I run the command in ngspice window.

Just insert the

plot @m.xm1.....[gm]

in the

.control

block. Does this not work?

Tried the above but same result. @Stefan Schippers

@Pranav Lulu you need to quote the node names since they contain square brackets. You can also plot gm vs id, like this: plot "@m.xm2.msky130_fd_pr__nfet_01v8_lvt[gm]" vs "@m.xm2.msky130_fd_pr__nfet_01v8_lvt[id]"

Hi @Pranav Lulu @Stefan Schippers, I also have the same problem, even with the gm terms inside quotation marks. The graph only showed when the plot command is called from inside the newly popped up terminal

@Stefan Schippers Tried using the quotes but still no change

Hi @Pranav Lulu, I think I found out the problem, so we need to add run after .control

Hey!! @Edo Jayakusuma. It worked!!

either use a run command or start ngspice with the -a option (autorun)

@Stefan Schippers How can

gm

of nfet be plotted in xschem internal waveform viewer ?

gm can be plotted in xschem if the gm is saved in the raw file. you must save gm nodes explicitly:

.save
+ @m.xm1.msky130_fd_pr__nfet_01v8_lvt[gm]
+ @m.xm2.msky130_fd_pr__nfet_01v8_lvt[gm]
+ @m.xm4.msky130_fd_pr__nfet_01v8_lvt[gm]
+ @m.xm3.msky130_fd_pr__pfet_01v8_lvt[gm]
+ @m.xm5.msky130_fd_pr__pfet_01v8_lvt[gm]
+ v(@m.xm5.msky130_fd_pr__pfet_01v8_lvt[vth])

Then these nodes will appear in the graph dialog box. Double click on the list on the left to add it in the graph.

Thanks, got the point. Able to plot

gm

now.

What are y'all using for gm and gm/Id simulations? Might be that I have, at considerable length, re-invented the wheel: https://github.com/tatzelbrumm/sky130_cm_ip__biasgen/blob/main/xschem/nfet_g5v0d10v5_gmtest.sch https://github.com/tatzelbrumm/sky130_cm_ip__biasgen/blob/main/xschem/nfet_g5v0d10v5_gotest.sch https://github.com/tatzelbrumm/sky130_cm_ip__biasgen/blob/main/xschem/pfet_g5v0d10v5_gmtest.sch https://github.com/tatzelbrumm/sky130_cm_ip__biasgen/blob/main/xschem/pfet_g5v0d10v5_gotest.sch but in any case, with a some silicon still promised after the April 2024 tapeout, how could that silicon area best be utilized not as proof-of-concept for using CACE to document specifications that look suitably presentable to suits, but are based on unreliable models, but instead to put structures on silicon that would improve the underlying models?

@Christoph Maier No I don't know about this variable. Is it a tring type? I dont see a

region

parameter in the whole models_fet.spice in the sky130 models.