Hello everyone! I'm new to electronics, and frustrated by a recent test-circuit setup...
I'm measuring the frequency response of an oscillator. I use Dynamic signal analyzer to provide with AC actuation (and the same actuation source connects to analyzer reference channel), and lock-in amplifier for demodulation, filter, etc.. The lock-in amplifier picks up oscillator's output, processes it and sends it to DSA which plots the Bode plot.
The magnitude-frequency curve is correct. It's easy to locate the resonant frequency on the curve. But the phase-frequency curve is a mess -- it's just nothing more than random noise drifting up and down during the whole frequency range. I'm puzzled. How can the two parts of Bode plot vary? They should be consistent, shouldn't they?
-- Or there's something wrong with the lock-in amplifier? Anyone has an idea? I really appreciate any suggestions.

Thanks,
J

 

It would be helpful to see your test-setup (block diagram).

 

I interpret your results to mean that the phase response of the oscillator is moving around at random and is not a well defined function of frequency. I also think this analysis is not appropriate for oscillators in the same way it is appropriate for filters or amplifiers.