it seems like the pnpMPA model and pnpMPA.sym have different pin-orders

[olisnr]

in the file IHP-Open-PDK-dev/ihp-sg13g2/libs.tech/ngspice/models/sg13g2_hbt_mod.lib there is a model:

.subckt pnpMPA e b c
.param a=2p p=6u ac=13.33p pc=14.64u
+ dev_a=a*1e12 dev_p=p*1e6 sub_a=ac*1e12 sub_p=pc*1e6

QpnpMPA e b c pnpMPA_mod area=dev_a
...

but if i run the netlist-generation of Xschem for IHP-Open-PDK-dev/ihp-sg13g2/libs.tech/xschem/sg13g2_tests/dc_pnpMPA.sch it generates:

** sch_path: IHP-Open-PDK-dev/ihp-sg13g2/libs.tech/xschem/sg13g2_tests/dc_pnpMPA.sch
**.subckt dc_pnpMPA
XQ1 net3 net1 E1 pnpMPA
Ve net2 E1 0
.save i(ve)
Vb net1 GND 0
.save i(vb)
Vc net3 GND 0
.save i(vc)
I0 GND net2 1u
...

it seems to me that c and e are interchanged somewhere.

[olisnr]

i think the description of the model is wrong and should be QpnpMPA c b e pnpMPA_mod area=dev_a because the gain of the transistor is a bit over 1, and thats like in the measurement data, where the gain is typical about 1.5. and if i flip the pnpMPA in the test-circuit so that c is e and versa vice, the gain is much lower.

i would also like to know, why the different DUT-measurements of the pnpMPA in IHP-Open-PDK-dev/ihp-sg13g2/libs.doc/meas/pnpMPA are that different. i dont see any differences in the headers of fg_vcb0_d0406_DUT1.mdm .. DUT17.mdm, its always

  TNOM "27"
  TEMP "27"
  TECHNO "SG13S"
  LOT "PQA701"
  WAFER "17"
  CHIP "d0406"
  MODULE "T323_pnpMPA"

so i would think, the data should be more or less the same for all DUTs

[sergeiandreyev]

the comments section has a specific DUT defined for each of the .mdm files, and these are different:

  DEV_NAME "DUT4"

[sergeiandreyev]

on the pin order, I'm checking internally

[olisnr]

thanks a lot!