RFICDUDE
|
If you have a perfect constant envelope signal then AM-PM probably doesn't matter much because there is no AM in the input signal. However, the modulator before a PA may introduce a small amount of AM due to the imperfect amplitude balance, phase balance, and nonlinearity in the modulator (this can convert PM-AM). If the input signal contains enough AM then the AM-PM of the amplifier will interact with the signal plus AM to generated more distortion at the output. The end result is spectral regrowth and worse phase error since now you get extra deviations in phase which are dependent on the instantaneous amplitude variation.
I don't think simulating AM-PM with PAC analysis is the 100% correct thing to do. You need the large signal results to obtain an accurate representation of the AM-PM characteristic. So, you might try measuring the AM-PM of the carrier signal from the PSS simulation. Just measure the large signal complex gain from input to output while sweeping carrier power.
Here are some references that may be helpful to you
Jones, A.E.; Wilkinson, T.A.H.; Gardiner, J.G., "Effects of modulator deficiencies and amplifier nonlinearities on the phase accuracy of GMSK signalling," Communications, Speech and Vision, IEE Proceedings I , vol.140, no.2, pp.157-162, Apr 1993
Hyunchol Shin; Walker, B.; Dongling Pan; Dunworth, J.; Jaffee, J., "Analysis of spectral spreading in a phase-modulated system for 1.75-GHz GSM RF transmitter design," Custom Integrated Circuits Conference, 2003. Proceedings of the IEEE 2003 , vol., no., pp. 103-106, 21-24 Sept. 2003
|