question on osillation mechanism of an amplifier

[abcyin]

Hi, all

sometimes, after you designed an amplifier, you find that the .tran response will show you an osillation phenomenon, maybe the phase margin is not enough for the amplifier, but is there any other mechanism for the osillation, can anybody give me some summarized conclusion for the mechanism.
and maybe there is some literature to explain this, so who can recommand some of them.

Thx in advance!

 

[pico]

Oscillation will occur when there is a feedback loop with unity gain and zero degree phase shift.

 

[pfd001]

loop gain is larger than 1 and phase shift is 360

 

[leohart]

How about loop gain>1 and phase shift=180?Seems we define this situation as unstable if we check the phase margin plot with loop gain=1 happens at phase shift>180.

But it really makes no different when we use A=A/(1+A*beta)

 

[m.ghafoori]

In practice, when the S parameters of the circuit does not satisfy the requirements, or in other words, the amplifier is not matched, it begins oscillation.
In order to avoid oscillation, the return signal should be attenuated at the output port of the amplifier. Therefore, you can use a 2-3 dB R-Attenuator after the decoupling capacitor in output port. If you need more information, let me know.

 

[Sadegh.j]

There are actually many many texts available on oscillation and of course many many papers, you may find some of them in:
https://www.its.caltech.edu/~hajimiri/publications.html

But to summarize, oscillation has two main reasons:

One is what you actually mentioned and that is the phase margin and stability of your amplifier.
If your substrate is not ideal, some time and in certain cases, it might act as your feedback which again cause instability.

Cheers