If I try to simulate a bjt astable multivibrator made with two identical npn, two identical base resistors, two identical collector resistors and two identical capacitors cross connecting the base and collector node of the two bjts powered with a 15 V battery through a SPST switch, I have no oscillations starting while closing the switch. Putting two oscilloscope probes on C and B nodes shows only noise. Putting voltage probes on the same nodes shows correct saturation voltages. Same way changing hfe of the bjt as they aren't quite identical. Oscillation starts adding and operating a momentary pushbutton between a B and E nodes, but is very erratic and damps off after a while. Adding a couple of diodes before the bjts bases to avoid base junction reverse avalanche does nothing. I suppose there is something in the bjt AC model that behaves uncorrectly in case of positive feedback. Simulating a single transistor
BF amplifier with stabilized polarization (R1,R2,Re)seems to give correct results.