Hi @Stefan Schippers, Do you know why some curves are cut and cannot be seen completely? It seems to be a bug, since it is not enough to select the curve, since only one curve is divided into 3 or 4 datasets, so I cannot quickly know what value varied with the dataset, do you have any solution for this?

If curve is not monotonically in the x direction try to manual zoom the x axis. For more info , please send a complete testcase (the .sch schematic and *.sym / *.sch of subschematics if any).

The problem is that I would like to identify the exact dataset associated with my parameter variation, but a single curve breaks into 3, I have seen it in several sweeps, but it only happens in that graph, it does not happen in the ngspice viewer

Thank you for providing the test case. Please update xschem, i have made a fix and now curves should not be broken anymore. In some cases selecting a curve with the 't' key is difficult if there are many sweeps. In this case do this: • Double click the graph • in the graph dialog box put a number in the dataset you want to see. The graph will display only that dataset. Please check your while loop. You are simulating two times with L=0.15 and not simulating L=3 (last dataset is L=2.95) Move the

alterparam L2 = $&l_act

line and the reset line before the

dc VSG1 0.7 -1.8 -0.01

line, the

run

instruction is not needed.

let start_l = 0.15
let stop_l = 3
let delta_l = 0.05
let l_act = start_l
reset
while l_act le stop_l
   alterparam L2 = $&l_act
   reset
   dc VSG1 0.7 -1.8 -0.01 
   *plot @M.XM2.msky130_fd_pr__pfet_01v8[id] ylabel Id xlabel Vds
   let gmidp=(@M.XM2.msky130_fd_pr__pfet_01v8[gm])/(@M.XM2.msky130_fd_pr__pfet_01v8[id])
   let IdWp=(@M.XM2.msky130_fd_pr__pfet_01v8[id])/(@M.XM2.msky130_fd_pr__pfet_01v8[w])
   let idp=@M.XM2.msky130_fd_pr__pfet_01v8[id]
   let gmp=@M.XM2.msky130_fd_pr__pfet_01v8[gm]
   let gdsp=@M.XM2.msky130_fd_pr__pfet_01v8[gds]
   let gmgdsp=gmp/gdsp
   let cggp=@M.XM2.msky130_fd_pr__pfet_01v8[cgg]
   let vovp=@M.XM2.msky130_fd_pr__pfet_01v8[vgs]-@M.XM2.msky130_fd_pr__pfet_01v8[vth]
   let ftp=gmp/(cggp*2*pi)
   let fup=gmp/((2e-12)*2*pi)
   *plot @M.XM2.msky130_fd_pr__pfet_01v8[gm]/@M.XM2.msky130_fd_pr__pfet_01v8[id]
   remzerovec
   save all
   set filetype = binary
   write pmos.raw 
   let l_act = l_act + delta_l
   set appendwrite
end

Hi, thanks for that, I managed to identify another error, or I don't know how to solve it, if I graph for example in that same gm/gds test, the datasets start from 0, but if now in sweep I put it to be graphed with respect to gm/id, the dataseet no longer are from 0, otherwise unknown values

What is the problem? I looked at graph gm/gds for the pmos and it has gmidp as the sweep variable, everything works here. What are the steps to reproduce the problem?