What are parameters of IP3 in ADS for a 2 tone modulation?

[f_t]

Hello friends,

I want to obtain IP3 of my PA in ADS. There is a simulation tool names IP3out which can calculate IP3, but there are two parameters in this tool which I do not know what they are. I have done a 2 tone modulation on my PA.




harmonic frequency indices for the fundamental frequency (fundFreq)
and
harmonic frequency indices for the intermodulation frequency (imFreq)

What are they?

 

[BigBoss]

It means...
When you apply two equal tone to the input of a device (amplifer,mixer etc.) the output will have the main carriers (tones) and their combinations due to nonlinear effect of the device.( see math. background of nonlinearity)
So,
fundFreq={1,0}=F1*1+F2*0=F1 or can be {0,1}=F1*0+F2*1=F2
Similarly
imFreq={2,-1}=F1*2-F1*1=2F2-F1 or can be such combinations {1,-2} so on..

This measurement block computes the harmonic power at those frequecies with load voltage over a load impedance (can be complex Z=x+-j*y) and IP3 equation that is

OIP3=P1+(P1-P{2,-1})/2

And this is valid only and only the main tone power level is at least 20dB lower than P1dB compression level.

is it clear ?

 

[f_t]

It means...
When you apply two equal tone to the input of a device (amplifer,mixer etc.) the output will have the main carriers (tones) and their combinations due to nonlinear effect of the device.( see math. background of nonlinearity)
So,
fundFreq={1,0}=F1*1+F2*0=F1 or can be {0,1}=F1*0+F2*1=F2
Similarly
imFreq={2,-1}=F1*2-F1*1=2F2-F1 or can be such combinations {1,-2} so on..

This measurement block computes the harmonic power at those frequecies with load voltage over a load impedance (can be complex Z=x+-j*y) and IP3 equation that is

OIP3=P1+(P1-P{2,-1})/2

And this is valid only and only the main tone power level is at least 20dB lower than P1dB compression level.

is it clear ?

Yes, It is cleasr. Thanks.
To calculate Input IP3 we should use the below equation. Am I right?
IP3INPUT = (IP3OUTPUT - Gain) {dBm}

If the condition for IP3 (20dB<P1dB compression) does not happen, both our calculation for IP3 and the value which ADS gives us will not valid? In this case what we should do to obtain IP3? I mean our circuit does not work well and we should change it or we should find IP3 by another way?

 

[RealAEL]

There is no other way that is accurate. The measurement/simulation has to be done with the device operating in a linear mode. The ADS notes say:

To measure the third-order intercept point, you must setup a Harmonic Balance simulation with the input signal driving the circuit in the linear range. Input power is typically set 10 dB below the 1 dB gain compression point. If you simulate the circuit in the nonlinear region, the calculated results will be incorrect.

The ip3_out measurement will give results if even if the input power is too high but the results will not be corrent. An power 10 dB below 1 dB gain compression may be a bit high. I agree with BigBoss that 20 dB is much better but that is maximum value, any value less than this will give a good result.

I usually do IP3 simulations with swept input power starting at low power levels. Provided the ip3_out results are flat (constant) with increasing input power then I know the data is being calculated correctly and that is the answer required. If not then the power likely too high.

 

[f_t]

There is no other way that is accurate. The measurement/simulation has to be done with the device operating in a linear mode. The ADS notes say:



The ip3_out measurement will give results if even if the input power is too high but the results will not be corrent. An power 10 dB below 1 dB gain compression may be a bit high. I agree with BigBoss that 20 dB is much better but that is maximum value, any value less than this will give a good result.

I usually do IP3 simulations with swept input power starting at low power levels. Provided the ip3_out results are flat (constant) with increasing input power then I know the data is being calculated correctly and that is the answer required. If not then the power likely too high.

Ok, just take a look on the figure of my circuit and its result. I have used a 2 tones signal modulation in input. Figure of output power vs input power is strange! Besides, ADS shows 36 dBm OIP3, but according to the figure OIP3 is 41 dBm. The amplifier is a class-E power amplifier which works at 2.3 GHz.

 

[vkschaudhary2010]

hello f_t,
i am also working on class-e power amplifier and now i am getting some problerm in calculating PAE(power added efficiency) and gain. and need some help for harmonic simulation and load pull simulation.
can you share your class e amplifier design with me, it will be a great help.
thank you

 

[f_t]

hello f_t,
i am also working on class-e power amplifier and now i am getting some problerm in calculating PAE(power added efficiency) and gain. and need some help for harmonic simulation and load pull simulation.
can you share your class e amplifier design with me, it will be a great help.
thank you

Hi,
Do you mean you need my simulations in ADS?

 

[vkschaudhary2010]

Hi f_t,
yes i need your simulation, it will be a great help. It gives me an idea about procedure in ADS.

 

[BigBoss]

Class-E amplifiers are strongly nonlinear amplifiers and essentially switch mode amplifiers.So, drain current looks a switched current ( ON/OFF) therefore OIP3 is meaningless for this class.OIP3 is valid for linear or mild nonlinear amplifiers.That's why your setup and simulation is not correct because you have to drive the Gate hardly and it will make OIP3 equation/definition invalid.

 

[f_t]

Hi f_t,
yes i need your simulation, it will be a great help. It gives me an idea about procedure in ADS.

Hi,

No problem. Please send your email.