Continue to Site

The importance of loop gain

Status
Not open for further replies.

sikhjumn

Newbie level 2
I'm confused about the usage of loop gain of Op Amp.
Can anyone tell me the importance of loop gain,e.g. open loop gain&close loop gain?

Thx a lot.

sikhjumn said:
I'm confused about the usage of loop gain of Op Amp.
Can anyone tell me the importance of loop gain,e.g. open loop gain&close loop gain?

Thx a lot.

Yes, sometimes one is confused, but it is important to know the difference.
There are three important gain parameters:

1.) open loop gain: gain of an amplifier without feedback (Aol)
2.) closed loop gain : gain of an amplifier with feedback (Acl<Aol)
3.) loop gain: gain of all components of the (open) loop. Example: opamp with resistive feedback: loop gain L= Aol*F (F=feedback factor R1/(R1+R2).
___________
The loop gain L is the most important parameter for stability investigations.

Maybe...Are there any practical example for loop gain?
Or any example that the loop gain directly affect the results?

While i search the loop gain through internet, the example always based on mathematics. However, I need more examples baesd on macro views, or any product based on the principle of loop gain.

sikhjumn said:
Maybe...Are there any practical example for loop gain?
Or any example that the loop gain directly affect the results?
While i search the loop gain through internet, the example always based on mathematics. However, I need more examples baesd on macro views, or any product based on the principle of loop gain.

* One typical example was given in my first answer
* based on the classical feedback scheme, Blacks formula shows the importance of the loop gain L(jw):
Acl=Aol/(1-Aol*F)=Aol/(1-L)
(Notation as in my first answer)
* The sign of L determines if you have positive or negative feedback, and dont forget: L=L(jw), which means that L decreases (and its phase changes!) with rising frequencies
* Special case: If there is one frequency with |L|=1 (0dB)
and phase of L is 360 deg, the circuit design for amplification will become unstable.
*Recommendation: Try to become familiar with stability criteria; the you will see the importance of loop gain.

Yes, as LvW said ,loop gain is most important in some case. such sc circuit for adc.
but, a litter confused me is the simple way of ac analysis for loop gain and phase?.since i've known how to do for Aol and Acl. can you help for this,LvW
ths a lot.

LvW said:
sikhjumn said:
Maybe...Are there any practical example for loop gain?
Or any example that the loop gain directly affect the results?
While i search the loop gain through internet, the example always based on mathematics. However, I need more examples baesd on macro views, or any product based on the principle of loop gain.

* One typical example was given in my first answer
* based on the classical feedback scheme, Blacks formula shows the importance of the loop gain L(jw):
Acl=Aol/(1-Aol*F)=Aol/(1-L)
(Notation as in my first answer)
* The sign of L determines if you have positive or negative feedback, and dont forget: L=L(jw), which means that L decreases (and its phase changes!) with rising frequencies
* Special case: If there is one frequency with |L|=1 (0dB)
and phase of L is 360 deg, the circuit design for amplification will become unstable.
*Recommendation: Try to become familiar with stability criteria; the you will see the importance of loop gain.

fencl said:
.............
but, a litter confused me is the simple way of ac analysis for loop gain and phase?.since i've known how to do for Aol and Acl. can you help for this,LvW
ths a lot.

You ask for a way to measure/simulate loop gain?
This subject has been discussed intensively in this forum before.
For example, have a look at "similar topics" at the bottom of this page.
LvW

LvW said:
fencl said:
............
You ask for a way to measure/simulate loop gain?
This subject has been discussed intensively in this forum before.
For example, have a look at "similar topics" at the bottom of this page.
LvW

oh, i've got it ,thx!

Hi, LvW,
since i follow the loop gain simulation way.i've got singal-end output gain&phase ,it works well. but when i try to the similar way for fullydiff-opamp. it become strange.
usually i needed is sc circuit in adc or fullydiff in pga.so i still can't use it for identification.
have you simulate for fullydiff before? can you share some experience?

Regards!

LvW said:
fencl said:
.............
but, a litter confused me is the simple way of ac analysis for loop gain and phase?.since i've known how to do for Aol and Acl. can you help for this,LvW
ths a lot.

You ask for a way to measure/simulate loop gain?
This subject has been discussed intensively in this forum before.
For example, have a look at "similar topics" at the bottom of this page.
LvW

Yes you can still use loopgain simulation methodology in cadence. However before running "stb" analysis you have to run transients till your CM is settled. An easier way is to use ideal CMFB circuit and then do can skip the transient analysis part.