Hello, when designing an amplifier with feedback its transmission function can cause a singularity in the denomintor at certaing frequency.
thus we could get infinit gain.
Why in real life the singularity of an amplifier causes it to be an oscilator?
Thanks.

 

How do you think it happens?

 

Hello nsaspook,
I dont understand either how the amplifier becomes an oscilator
if we have infinit gain then our signal becomes VDD and thats it.

 

High gain doesn't necessarily mean an amplifier will oscillate.

 

Hello nsaspook,
I dont understand either how the amplifier becomes an oscilator
if we have infinit gain then our signal becomes VDD and thats it.

Obviously you don't have infinite gain but what you do have is the electrical equivalent of a spinning flywheel at that singularity that stores and releases energy at the proper phase shift as a condition for oscillation.
https://www.allaboutcircuits.com/technical-articles/how-to-use-nyquist-plots-for-stability-analysis/

https://convexoptimization.com/TOOLS/Black.pdf
AVOID SINGING

 

Hello nsaspook,
I dont understand either how the amplifier becomes an oscilator
if we have infinit gain then our signal becomes VDD and thats it.

If the feedback varies with frequency, what happens to the phase?

 

If the output-to-input feedback becomes positive with a gain >1 due to phase-shift at a certain frequency then the amplifier is likely to oscillate at that frequency.

 

Hello crutschow,So what kind of feedback response we want for an amplifier?
amplifier suppose to amplify our input RF(sinus) singnal from 10dbm to 30dBm.(for example)
at some frequency our H turns to negative thus the numerator is posivitive.
So what is the problem in that?
We have a positive gain.Why negative H is a problem?
Thanks.

 

Generally, each device become to oscillating when fulfills the 3 factors simultaneously: 1) the input somehow gets the at least part of output signal back 2) the phase of this signal on the frequency of generation is perfectly synphase 3) the gain on this frequency is at least 1.00 or higher. Thus, generation is more frequently possible in practice as avoided.

 

In practice, an amplifier that shouldn't oscillate certainly will and an oscillator that should oscillate certainly won't!