I'm playing around with a circuit to drive a LED based on the output from a current sensing transformer. The idea is to have the LED appear proportionally brighter with the output from the current sensing transformer. I've figured out I need an antilog amplifier for that, but since the output current is too small to drive the LED, I've added a PNP buffer after it. Simulating this in SPICE seems to work well for DC, but trying to run the simulation in AC (as the sensing transformer should work on mains), having added some components to smooth out the input to the antilog amp, the simulation never finishes, so I'm assuming oscillation.
Closest I got was this:
If I increase R1 any further or even R5, it stops working properly, but I do not really understand why. Can anyone explain to me why the AC version is not working properly?
DC only sim looks good:
