ip3 measurement
Thank you
I think it is about the BW of you baseband signal. If you channel is 30 kHz , it should be about 30 KHz.
The frequency separation of the two input carriers shall be selected so that the 3rd order IM products lies within the passband as well.
In the case of a 30 kHz BW, use a maximum of 10 kHz frequency separation. At 10 kHz frequency separation, the measured IM products will have a frequency separation of 30 kHz, which will exactly fit within the 30 kHz BW.
I think within the BW, select the frequency separation as small as possiable, two large will make the measurment not accuracy, maybe 5 K is suitable
Usually, IP3 performances of an amplifier (or in general of an apparatus) change very little with frequency.
So it's not the channel separation that should set the frequency separation but the knowing that you are into the bandwith of your apparatus. Frequencies should be far enough to be seen cleany (small phase noise) into Spectrum Analizer but closer enough to see 3rd Order and 5th order into the BW and into the SA screen.
A note about power level required.
To be sure enough that the 3rd Order will be seen far enough from noise but the signal still remain in linear region, as rule of thumb, is to tune the level to see
the 3rd Order 40 dB smaller than test signal.
A note about accuracy.
IP3 measurement give not a number, give a band of acceptance. Looking for high accuracy is not required and make a non sense.
measurement 相关文章:
