Suggestions on methods of IP2 measurement

[RCinFLA]

What defines the correct frequency of IIP2 measurment?

Where the measurement is not effected by selectivity in the circuit.

 

[morray]

Do you mean selectivity which defined by the bandpass filter prior the A2D?

 

[RCinFLA]

Yes, if that is the application.

IP2 is cross modulation. Just about all active devices have some gain compression with increasing signal levels. I am not just talking about high end clipping. The gain compression causes unsymetrical waveform distortion, like when the top half of a sinewave amplitude is not as great as the lower half of sinewave.

Yes, the Fourier transform of a non symetrical waveform has even harmonics but in the case of intermodulation, the presences of a large interferer, along with the active devices gain compression, causes amplitude modulation to be imparted on any smaller signal that is riding up and down the gain compression slope caused by the larger interferer signal swing.

Any large interferer with a modulation envelope that has high peak to minimum ratio will impart its modulation on a lesser signal when passed through an active device with gain compression.

Take cellular for example, GSM GMSK voice as almost no amplitude variation. EDGE modulation does. CDMA modulation does. Muli-level QAM does. The higher the peak to average envelope power the more damaging. OFDM has very high peak to average envelope ratio.

This is what you are trying to quantify with IP2 specification, not how much second harmonic is produced.

Two tones is an easy way to create an interferer with a high peak to minimum amplitude ratio.

If the two CW tones have 0 db difference the peak to minimum ratio approaches infinity.

0.1 db delta in two tone amplitude has 45 db peak to min ratio. (0.1 db is about the best expected accuracy of setting two generators amplitude)

1.0 db delta of two tones yields a 25 db peak to min ratio.
3.0 db delta of two tones yields a 15 db peak to min ratio.
6.0 db delta of two tones yields a 9.6 db peak to min ratio.

 

[varuniyengar]

Design ckt on ADS and find IP2 using inbuilt template

 

[morray]

Thank you guys for your answers. My question is more a matter of definition, not practical. When someone sais: IIP2 @ 1000MHz, does he mean that the intermod is located at that freq, or that the two tones are in that freq. The way I see it, it means that the intermod should fall in that freq, since the idea behind the IP2 measurment is to tell how much sensitivity is degraded due to the intermod at the same frequency. And then the question rises again - What should be the freqeuncy of the two tones used to create the intermod? What do you say?
Thanks again.

 

[varuniyengar]

IIP2 means the point at which the Fundamental and Second Harmonic Voltages at the output of a non linear device meet with respect to change in input voltage.