# regarding negative feedback

Discussion in 'General Electronics Chat' started by elisettynarendra, Oct 10, 2010.

1. ### elisettynarendra Thread Starter Member

May 17, 2010
42
1

In the picture , the output 'eo' corresponds to input Av*(eg-Mv*eo)..but initially we don't have any feedback signal..
so at first the output will be (Av*eg) which is not 'eo' say 'x' .now this is taken as feedback and them amplified say ( Mv*x).
now the output corresponds to Av*( eg-Mv*x )..so the output decreases..and the magnitude of feedback signal also decreases ,so the input increases and output increases ..so as feedback signal increase ..and then o/p decrease..
My doubt is that if this continues in circular fashion then how can we get constant o/p 'eo' as represented in the picture.
plzz help ..

Last edited: Oct 10, 2010
2. ### eblc1388 AAC Fanatic!

Nov 28, 2008
1,543
104
You're right. This is where someone earn their wages.

A constant output is a particular or desired condition of the feedback system rather than the norm.

Without careful design the output of a feedback system will not be stable.

As an engineer or designer, your task is to make sure it remains stable and not oscillatory.

How? I don't know as some of my feedback circuits are oscillating like mad.

Last edited: Oct 10, 2010
3. ### elisettynarendra Thread Starter Member

May 17, 2010
42
1
i think these oscillations happens in each n every case.. then how can we take measures to keep output constant..i referred as many as 10 e-books but everywhere i found the o/p taken constant..none discussed abt these variations in o/p.. does this happen really or am i wrong in analyzing..

4. ### The Electrician AAC Fanatic!

Oct 9, 2007
2,621
468
The important question is whether or not these oscillations die out with passing time, or whether they continue to increase in amplitude until the amplifier saturates.

Your job as a designer is to use tools such as Bode plots, the Nyquist criterion, etc., to determine if the oscillations die out. If they do, the system is stable.

In many cases, any oscillations are undesirable, but sometimes a few cycles of oscillation can be tolerated if they die out quickly. The closed loop step response shows these oscillations if they are present. Often, it is desired that there be no oscillations, or at most one. The amount of "damping" in the system controls the oscillations. A system that is overdamped will approach the final state with no "overshoot".

Such oscillations are sometimes called "ringing" if the system is "underdamped". The oscillations that occur with "underdamped" systems are an indicator that the system is close to "instability".

You might look for discussion of the words I've put in quotes, in connection with feedback systems.

Also search for explanations of the terms "gain margin" and "phase margin".

5. ### elisettynarendra Thread Starter Member

May 17, 2010
42
1
thanks for your relpy..i will go through .