does IIP3 vary with the amplititude of input two tone?

the method of PSS+PAC was used to simulate the IIP
and from the simulaiton ,i got the iip3 varied with the amplitude of input signals.
but in my opioion ,the linearity of system can't swing with the input signal.
do i have a mistake? can anybody tell me the truth?

yes
linearity swings with the input signal.
regards

but the IIP3 of system must be not influence by the amplitude of input signal.
and in my simulaiton, input signal is small

IIP3(dBm) = [Pout(dBm) + IM3(dBc) / 2] ? Gain(dB)

But this equation is valid only when your amplifier is not in compression. When you are in compression (high input level) the Gain is changing.

Does it mean that the IIP3 should be equal when the input signal small enought?
complex filter and AGC is measure in my work, the IIP3 of AGC is independ of the input signal basically, but the result of complex filter depend on input signal power.

Hi
sorry, I think IIP3 does not change with input signal power because we assume that it is linear.
I think you don't measure IIP3 correctly.
regards

my testbench is as follow:
input signal source is a Port with a large sine signal and a PAC signal, they are same amplitude --45dbm, the frequence of sine signal is 2.1Mhz.
PSS analysis was done first ,then a single-point PAC is done,the frequence point is 2.2Mhz.

analysis('pac ?start "2.2M" ?stop "" ?maxsideband "5" )
analysis('Pss ?fund "1M" ?harms "10" ?errpreset "moderate" ?tstab "10u" ?saveinit
"yes" ?method "gear2only" ?tstabmethod "gear2only" ?relref "sigglobal ")

2Mhz was choosen as 3-rd harmonics and 2.2M was choosen as fundamental when IP3 was plotted at the end.

Hi
I don't know that it is true.
see
http://www.ek.isy.liu.se/courses/tse...A_Tutorial.pdf
http://www.ek.isy.liu.se/courses/tse...r_Tutorial.pdf
regards

You need to make sure that
1- you are operating well below the compression point (Determined in a seperate simulation)
2-You don't have any 5th order harmonics starting to come up. The 3-2 fifth order products will show up at the same frequency as the third order products.

Dave

The p1dbm of my circuit is 1.8dbm, -45dbm input signal was set in ip3 simulation.

Can you tell me more detail? How can i meke sure 5th order harmonics not to come up?

Thanks

Even below ICP, the gain starte to decrease. So IIP3 is dependent on which point of power you extrapolate. The lower power you extraploate, the more accurate IIP3 you will get, theoretically.
However, if the power is too low, noise and higher order products starte to be evident.

You can make sure that there are no 5th order products by looking at 2 * delta f offset from the 2 input tones. This is the first place that the 5th order products will show up seperatly from the 3rd order products.

Dave

I agree with hr_rezaee.

your measurement is maybe error. the IIP3 must be a value when the signal isn't compression.

So you only measure the cross of the linear line and actual line.you will be care of the linear line must descript by your point which is set in PSS or PAC.

At last, you can obtain a correct result.

Best Regards!