@Stefan Schippers I tried simulating the 5v5 npn in xschem and it crashes with the following error:

Netlist line no. 0:
Undefined parameter [dkisnpn1x2]
Netlist line no. 0:
 Expression err: 7.98e-018*dkisnpn1x2+mc_mm_switch*agauss(0,1.0,1)*(7.98e-018*dkisnpn1x2*0.0036/sqrt(mult))} subs=1 rb=885.61 re=15.0 irb=4.424e-005 rc=61.677 rbm=256.08 bf={37.54*dkbfnpn1x2+mc_mm_switch*agauss(0,1.0,1)*(37.75*dkbfnpn1x2*0.0465/sqrt(mult))} nf=1.0405 vaf=100 ikf=0.0011462 ise=5.77e-016 ne=1.7924 ns=1.0 br=1.0 ibc=0 iss=0 nr=0.96012 var=0 ikr=4.032e-008 nkf=0.5 isc=0.0 xtb=0 xti=1.0713 eg=1.188 gap1=0.0 gap2=0 ctc=0.0 cte=0 cts=0 tlev=0 tlevc=0.0 tvjc=0 tvje=0 tvjs=0.0 tis1=0 tise1=0 tisc1=0.0 tnf1=4.208e-005 tnr1=-0.000522 tne1=0 tnc1=0.0 tbf1=0.00776 tbr1=0 tiss1=0 tvaf1=0.0 tvar1=0 tikf1=-0.0074 tikr1=0 tns1=0.0 trb1=0 trc1=0 tre1=0 tirb1=0.0 trm1=0 tmjc1=0 tmje1=0 tmjs1=0.0 ttf1=0 titf1=0 ttr1=0 tis2=4.0e-12 tise2=0 tisc2=0 tnf2=-3.372e-007 tnr2=1.8e-6 tne2=0 tnc2=0 tbf2=6.48e-006 tbr2=0.0 tiss2=0 tvaf2=0 tvar2=0 tikf2=4.0e-5 tikr2=0 tns2=0 trb2=0 trc2=0.0 tre2=0 tirb2=0 trm2=0 tmjc2=0.0 tmje2=0 tmjs2=0 ttf2=0 titf2=0.0
Netlist line no. 0:
Cannot compute substitute

The circuit is based on the test_bipolar.sch - which seems to run fine with the PNP devices, but not with the NPNs. I just swapped the device and made the appropriate changes (swapped the direction of the emitter current source, etc)

Can you post a simple simulation netlist? I expect that the issue is within the "include" statements in sky130.lib.spice or one of the files downstream. The "1x2" in the parameter name implies that it goes with one of the fixed-layout NPN devices. The original PDK conversion to open source made a huge mess of the BJT devices, and it has been slow work writing patch scripts to fix it all in open_pdks.

There it is. Slack didn't warn me about this thread for some reason..

** sch_path: /home/leonardo/proj_sky130/test_npn/schematic/test_npn.sch
**.subckt test_npn
XQ1 net2 net1 GND GND sky130_fd_pr__npn_05v5_w1p00l1p00
V1 net2 GND 3
.save i(v1)
I0 GND net1 1n
**** begin user architecture code



.op




.param mc_mm_switch=0
.param mc_pr_switch=0
.include /home/leonardo/pdk_files/open_pdks/sky130/sky130A/libs.tech/ngspice/corners/tt.spice
.include /home/leonardo/pdk_files/open_pdks/sky130/sky130A/libs.tech/ngspice/r+c/res_typical__cap_typical.spice
.include /home/leonardo/pdk_files/open_pdks/sky130/sky130A/libs.tech/ngspice/r+c/res_typical__cap_typical__lin.spice
.include /home/leonardo/pdk_files/open_pdks/sky130/sky130A/libs.tech/ngspice/corners/tt/specialized_cells.spice

**** end user architecture code
**.ends
.GLOBAL GND
.end

I can confirm the problem.

I'll keep it in mind. Thank you again!

yes the npn bipolars have some issues in sky130. To plot all device currents add

.option savecurrents

before running the analysis. Specific model parameters (like gm, vth, and so on) mus te saved explicitly, like for example:

save @m.x1.xm5.msky130_fd_pr__nfet_01v8[gm]
save @m.x1.xm5.msky130_fd_pr__nfet_01v8[gds]

For details see chapter 31 of the ngspice user manual