I am completely mystified by this thread so far. If I read things aright,
the present incarnation of the circuit contains no feedback, but is "very
unstable". I don't see how that can posibly be. Without feedback, there
can be no stability (as I understand the word) or instability.
Also I can't figure out whether the goal is to
a) debug the present incarnation of the circuit, or
b) just get something that works.
Note (a) may or may not be the easiest way to get to (b). If (a) is even
part of the goal, the discussion is going nowhere until we see a detailed
schematic of the circuit.