Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Explain what is the gain compression?

Status
Not open for further replies.

ziqbal

Newbie level 2
Joined
Mar 25, 2004
Messages
2
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
18
gain compression

I read about gain compression concept but still couldn't understand a single thing ... I am very much appreciated if somebody can briefly explain to me on gain compression....

appreciate :twisted:
 

Re: gain compression

In large signal amplifiers the output voltage swing cannot go higher than the power supply for example. As you start off with a low signal level and raise it the output at first rises and then after a while stays the same. If you plot the output verses the input you get a graph that is a straight line at the low signal levels and then bends over.

If you draw a straight line that is tangent to the small signal level region of the curve it will lie on the original curve at low levels and diverge at high levels.

In the region that they diverge, you can calculate the actual gain as out/in and this value is smaller than at low levels.
 

Re: gain compression

the gain is not the same as the signal becoming larger because of the non-learner,
when the sinal is large to some level the gain is becoming small and the compression come\\
 

gain compression

As the input signal increases for a practical two-port network, the output signal increases linearly until distortion product energy substantially combines with the fundamental output power. The result is that distortion products, along with fundamental, are observed at the output. The linear output power change versus an input power change no longer exists.
 

Re: gain compression

hi...

usually all amplifiers have the input vs output reltaion as below

y= a1*x + a2*x^2 + a3*x^3 + ..... (1)
where y is ofcourse o/p and x is i/p.

"Usually" a2 is quite small....and a4, a5, a6,....are also quite small.

example: for diffrential amplfiers a2=0

also a3 is negative....you can plot the o/p vs i/p transfer curve for non-linear amplifiers and u will see that the "magnitude" of slope of the curve , which is magnitude of gain, decreases....

such an o/p vs i/p curve can be expressed as equation (1) with a3<0

so gain= dy/dx = a1 + 3*a3*x^2

so u can clearly see that the gain actually depends on the amplitude of the input signal. and since a3<0 => gain decreases with large amplitude input "sine waves".

This is called gain compression characteristic of a non-linear amplifier.

do reply if u want more information....
 

Re: gain compression

Hi,

one more time using slightly different words.

Let us assume we have the perfectly linear amplifier with gain of G0, input signal x and output signal X=x+G0

Amplifier is supposed to be linear. In means that if you increase the input signal by amount of dx dB, the output signal will increase by the same amount dx dB, so the output will be X+dx. The gain is still G0.

We can continue to increase input level, the result will be the same up to the point where we start to drive the amplifier into the saturation. Close to the saturation, if we increase the input level by dx, the output signal will increase by dx1, dx1<dx and the gain of the amplifier at that point G1 will be less than G0.

We can continue this experiment. When the difference G1-G0=1dB, the amplifier is said to reach its 1-dB compression point at that input level.

We can continue this experiment further and after some input level, the increase in input level will cause no increase in output level. The amplifier is said to be saturated at that input level.

flyhigh
 

Re: gain compression

Another way to look at the non-linearity is by doing the Taylor expansion with respect to the bias point:

f(x)=f(x0) + f'(x0)(x-x0) + f''(x0)(x-x0)^2/2+....

That is why at large enough higher order terms your small signal analysis will NO LONGER VALID

T Zul
 

Re: gain compression

IN SIMPLE TERMS REAL DEVICES HAVE LIMITATION ON OUTPUT POWER DUE TO SATURATION, BELOW THAT LEVEL AT SMALL SIGNAL LEVELS OUTPUT IS SAY, 25*INPUT.
AS THE INPUT INCREASES OUTPUT ALSO INCREASES BUT AS THE SATURATION LEVEL REACHES OUTPUT CANNOT INCREASE PROPORTIONAL TO INPUT. IT IS SAID TO HAVE COMPRESSED. IT IS OBVIOUS THAT OUTPUT IS SAY, 24*INPUT.
HERE GAIN IS SAID TO GET COMPRESSED BY 24/ 25. THE RATIO IS EXPRESSED IN DB, GENERALLY.
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top