open-source-silicon.dev

I've been having trouble finding examples with ngspice since most people are using virtuoso which I believe has an option to calculate the needed plots workout writing the spice code

this is what I was  using on ngspice/octave to characterize the transistors on other processes
<https://gist.github.com/20Mhz/ab7c58c01b9e7259f8d1140a742e400a>
<https://gist.github.com/20Mhz/67d6b1b0e3b1f64e04837f54fd02d095>

I was thinking to get it updated to python…to be started…

my problem with ngspice is that save /plot/probe @mxx[gm] gives me a constant value regardless of any biasing conditions of the mos

well seems bogus, below used to work:
```vdsn     vdn 0         dc 'ds'
vgsn     vgn 0         dc 'gs'
vbsn     vbn 0         dc '-subvol'
mn       vdn vgn 0 vbn ami06n L='length' W='width'
vdsp     vdp 0         dc '-ds'
vgsp     vgp 0         dc '-gs'
vbsp     vbp 0         dc 'subvol'
mp       vdp vgp 0 vbp ami06p L='length' W='width'

.options dccap post brief accurate nomod
.op
.dc vdsn 0 'gsmax' 'gsstep' vgsn 0 'dsmax' 'dsstep'

.probe @mn[id]
.probe @mn[vth]
.probe @mn[gm]
.probe @mn[gmbs]```
are you doing something different?  I did have the wrapper on octave so each op is its own instance

9.png

*UPDATE* got it working now, dunno what i did wrong before...

.dc vg 0 1.8 0.01
.probe @m.xm2.msky130_fd_pr__nfet_01v8[gm]

What did you use for a circuit? Vg source direct to gate, 1.8v direct to d, ground s? Or do you have any other components

this example has drain fixed to 1.8 and varying gate

and you can plot vs id:
 plot @m.xm2.msky130_fd_pr__nfet_01v8[gm] vs i(vd2)

just vd=1.8 gate = parameter, source, body at 0. no other elements

last part gm starts to (slowly) decrease :slightly_smiling_face:

i put a 0V voltage source in series to the drain to get the current with a small expression ( i(vd2) ) instead of the long ngspice syntax for getting Id