Small signal VS. large signal analysis

[Octago]

Normally when we analyse a circuit, the small signal model is always used to explain the input impedance, gain...etc

But actually the input signal of a circuit is often relatively large,therefore it is not a "small signal" anymore.

At this time could we still use small signal model to explain the behavior of the circuit? If not, the large signal analysis which is about the DC operating point and so on can not be used to explain the complete circuit behavior?

So I am wondering when is the small signal analysis applicable? when the large signal? and in which case should I use small signal analysis and large signal analyis respectively?

 

[LvW]

Good question !

Perhaps, the answer will not satisfiy you, but there isn´t another:

There is no certain limit between small and large signal behaviour.
You have to specify the limit by yourself - depending on the error you are allowing.

To be very accurate, already an input signal of several millivolts at the base of a BJT is not compatible with small signal formulas - hoewever, the error will be negligible.

A practical method to check if you are in small or large signal operation would be to use a sinusoidal signal at the input of the amplifier and to analyze the THD of the output signal. As said before, the THD threshold is to be defined by yourself.
Hope this clarifiees a bit.

 

[mberesin]

Hi,
As far as I understand the problem small signal mode is used in dc and ac analysis.
Large signal models are used in tran analysis.

For example when You simulate amplifier ac characteristics You can apply 1V to the input and have thousands of Volts at output, what isn't realistic at all. This is the reason of using small signal model.
If You try to do tran simulation with input amplitude of 1 V the amplifier will surly saturate and the output amplitude can be maximum supply voltage.

Hope this could help.
Marcin

 

[ashish_chauhan]

I think its not like that you cant use large signal analysis all the time ... its just that most of the times the calculations involved go beyond what we can handle...
and we turn towards small signal analysis to get an aproximate picture.

Else... large signal is the real picture ( probably thats why many people say transient is god!)

 

[antonio_eda]

Small signal analysis is just a linear aproximation of the circuit behavior with ac signals. It is more or less accurate as write LvW and is really convenient tool in hand analysis.

 

[LvW]

Hi, octago: Because of mberesin´s answer it is necessary to ask you if your question is a general one or do you refer only to circuit simulation programs ?

 

[Octago]

My question is just a general question


quote="LvW"]Hi, octago: Because of mberesin´s answer it is necessary to ask you if your question is a general one or do you refer only to circuit simulation programs ?[/quote]

 

[edge_tv]

I thinke samll signal analysis is very usefull tool, also outside your operating point. But if you do that you have to know that your samll signal parameters like gm and Ro will change, and you have to understand why and how.
For Example, if I want to know what Gain I have at my Output swing border, I put the transistor in that condition look at the small signal paramters and check the gain. You will recognise that your gain is going down when you reach the upper or lower limet of your Supply Voltage (that's because your Ro goes down)
So I like it more to think in samll signal model, because it is valid in all regions but you have to know your samll signal parametres are not constant