I’m simulating with a builtin transistor model from KiCad 7, doing a DC sweep into the base of the transistor.
The voltage at the base looks legit, it hovers around the 0.7 volt from the B/E drop that I’d expect - but the voltage at the load (collector) remains unaffected. I am expecting this to be pulled down near ground, passing a current through the load.
I’ve spent hours messing around with these stupid transistors and I’m just not getting what I expect. (I have seen the NPN do what I expect before, and I’ve been next level stuck with a PNP simulation that refused to add up, but now even the NPN simulation is on crack)
You have named your load device ‘LOAD1’. This is an inductor, with ohmic resistance 0.
Rename it to RLOAD1, and it becomes a resistor, and the voltage v(/LOAD) drops during the sweep.
Oh damn it. I havent tried yet but that is almost certainty it. I think im only just realizing how “unsmart” nspice actually is, and that KiCad is a a bit of pig on a lipstick.
It seems very easy to make schematics that simulate radically differently to what you though you designed, if you missconfigure the instructions that map to spice.
So ngspice should be smart enough to recognize a certain graphic as a resistor?
I don’t think there’s ANY simulator out there able to do that.
That being said, KiCAD has an irritating “non-feature”: If you change a component value or reference in a symbol, the Spice model doesn’t change with it. You’ll have to check the Spice part to be certain.