In this article a RF Amplifier Design is proposed.
It's written that (page 2):
In an RF transmission network, if the reflection coefficient of a certain port is greater than 1, positive feedback will be introduced which may cause the amplifier unstable. In order to ensure unconditional stability, the real part of the impedance of the signal source, the load, the amplifier input, and the amplifier output, respectively, should be greater than 0.
To understand it, I think I should firstly understand what is meant for stability. Is it BIBO stability (Bounded Input Bounded Output) or does it mean that the amplifier starts oscillating?
The author speaks about load (or input) resistance and positive feedback. If the first one is negative, the last one occurs. Why? If the author refers to absolute negative resistance, it means that the latter changes its voltage proportionally to the current (like a positive resistor) but adds it to the input voltage (instead to subtract). I can't see what this has to do with positive feedback.
Positive feedback doesn't necessarily mean instability. It depends on the poles of the resulting transfer function. So, which is the point of this analysis?