[SOLVED] operational amplifier non linear effect

Hello guys

in the CMOS operational amplifier, we design it in all the transistors to be in sat region and we consider it as the linear region of amplifier operation. then after that we say this op-amp has non linearity problem because of the non linear relationship between the voltages and current in the cmos transistor and this what I am got confused.
in the second part of my question, in the very high gain Op-amp (>=80db) the range of the differental voltage is very very small in terms if several mV, so how like this range is suffering from the nonlinearity ??

Thank you all previously for your coming participation. and please if you have an articles it will be very helpful

senan said:

Hello guys

in the CMOS operational amplifier, we design it in all the transistors to be in sat region and we consider it as the linear region of amplifier operation. then after that we say this op-amp has non linearity problem because of the non linear relationship between the voltages and current in the cmos transistor and this what I am got confused.
in the second part of my question, in the very high gain Op-amp (>=80db) the range of the differental voltage is very very small in terms if several mV, so how like this range is suffering from the nonlinearity ??

Click to expand...

....we say...?
I think, the only remarkable non-linearity is caused by the limited slew rate of the unit.
Of course, strictly speaking, there is no real linear transfer characteristic at all - however, the applied negative feedback further reduces the (small) non-linearity of the open-loop gain.
Do you really think this is a problem?

Ok, then what mean of amplifier linearity?

senan said:

Ok, then what mean of amplifier linearity?

Click to expand...

Senan, if you really want to go into details (sources and amount of non-linearities) i recommend to you an application note from the very famous Barry Gilbert:
**broken link removed**

I have difficulties to hear a clear question from your post. If you are collecting material for a presentation or a thesis, it would be reasonable to review the known literature about the topic, e.g. the Nonlinearity and Mismatch chapter of Razavis Design of Analog ICs.

Linearity requirements can be quite different depending on the application, also frequency range or available loop gain to fight existing non-linearity. There are also nonlinear effects specific to particular amplifier topologies, e.g. crossover distortions.

I am sorry if my posts are not so clear

but I can repeat it again.

I am trying to write about the non linearity effect of the operational amplifier, I know that S.R is non linear phenomenon that can cause non linear distortion,but that is still considered as a large signal distortion at which the transistors not in the saturation.

then I confused when I read the attached papers and the page from Sedra smith book, they are talking about the non linearity effect wit the small signal condition, that confuse me more when I have very high gain so with very small differential voltage signal how can this non linearity ??



I believe that the MOS transistors are non linear but we are considering them linear by the condition of small siganl condition

FvM said:

I have difficulties to hear a clear question from your post. If you are collecting material for a presentation or a thesis, it would be reasonable to review the known literature about the topic, e.g. the Nonlinearity and Mismatch chapter of Razavis Design of Analog ICs.

Linearity requirements can be quite different depending on the application, also frequency range or available loop gain to fight existing non-linearity. There are also nonlinear effects specific to particular amplifier topologies, e.g. crossover distortions.

Click to expand...

Although the linked paper mentions OPs in the introducing "motivation" words, it doesn't particularly refer to OP design rather than circuits involving differential amplifiers with larger input voltage range.

But on the other hand (as already mentioned in my previous post), also OP circuits easily reach an operation range where the input differential voltage is 100 mV and more, introducing considerable diff-pair distortions, in other words leaving the small-signal domain.

For a well-defined application, the problems can be theoretically analysed or accessed in circuit simulation. Talking about linearity in a broad sense doesn't bring you far.

Quote senan: I believe that the MOS transistors are non linear but we are considering them linear by the condition of small siganl condition

Sena - exactly this is the point. Why you feel confused?
As I have mentioned already - there is no transfer characteristic that really is linear by 100%.
But - depending on the specific application - we accept (we have to accept!) a small portion of non-linearity.
That means - we do not care and consider the unit as linear. That`s all. However, of course we have to prove in advance if this simplification is allowed.
For example, this is evident for the mentioned case of slew rate influence.
By the way - that is the normal procedure in electronics. Take for example the part we call "resistor".
Is it an ohmic device with 100% accuracy? No - it is not. It has capacitive and inductive components.
But we don`t care as long as these parasitic influences are below a certain application-specific limit.
Also the math tells us that the tanh-characteristics you have shown in your reply certainly is not linear (tanh is not a linear function - even within a range of +- 1mV)

thank you guys for replying, I hope that you appreciate that I am beginner in this field , see you next with another basic questions