Designer’s Guide Community

NKS

New Member

Offline

Posts: 7
Seoul, Korea

Proper power settings for modulated RF signals-PAC
Jan 07^th, 2007, 10:06pm

Hello,

I am analyzing the distortion effects of modulated RF signal (GSM signal to be exact) on a non-linear circuit using PSS-PAC simulation but wondering whether my simulation settings are correct or not. For example, if the GMSK modulated RF signal of 30dBm power is centered around 900MHz, what power levels should I use to simulate the absolute levels of demodulated signals using the PAC "Modulated(PM)" feature? Currently, I am setting the power levels of both Amplitue1 and PAC magnitude parameters of Port signal source to
30dBm. But my question is, is this correct? I am concerned that I may be using power of twice the proper value.

Also, since varying the PAC magnitue value results in the demodulated signal levels of corresponding offsets, I am assuming that the PAC results should not to be interpreted as just conversion gains but
rather actual distortion levels. Is this correct?

Finally, is there a better way to represent GSM RF signal source for PSS-PAC simulation?

Best regards,

NKS

Back to top

IP Logged

Ken Kundert

Global Moderator

Offline

Posts: 2386
Silicon Valley

Re: Proper power settings for modulated RF signals
Reply #1 - Jan 8^th, 2007, 9:22am

PAC is a small-signal analysis, meaning that it operates on a linearized version of the circuit. Doubling the size of the stimulus will double the size of the response measured by PAC regardless of how large or small the input signal is.

The bottom line is that PAC measures transfer functions, not distortion. Having said that there is a way of measuring the intercept point of a circuit using PSS & PAC where the nonlinearity is excited in PSS and then measured using PAC. You can find out about that by reading http://www.designers-guide.org/Analysis/intercept-point.pdf.

-Ken

Back to top

IP Logged

NKS

New Member

Offline

Posts: 7
Seoul, Korea

Re: Proper power settings for modulated RF signals
Reply #2 - Jan 8^th, 2007, 3:43pm

Thank you Ken for your explanation! :)

At first I thought the PAC simulation results reflected transfer functin (conversion gain) but got confused when they varied with input levels; my logic was if they measured transfer function they should be independent of the input magnitude. Embarrassed

BTW, why don't the results remain constant?

I should have stated my motivation for using PSS-PAC in my previous posting. I am trying to use the tool to analyze the RF immunity characteristics of targeted circuits by measuring the resulting distortions when large RF signals with different types of modulation are applied. I thought conducting IIP3/OIP3 simulations would not distinguish between the AM and PM types of the source (RF) signal, and thus, somehow, not accurate.

Should I not worry about the types of moduation when the blocker signal is the modulated signal(RF) itself? That is, in your opinion, would IP3 simulations, which in this case would be treating the modulated contents as blocker signals, be the best way to analyze the RF immunity characteristics (measuring distortions-levels, waveforms) of target circuit? Other than transisent simulation(which takes too much time), of course. I would really appreciate your opinion or advice.

Best regards,

NKS

Back to top

IP Logged

Ken Kundert

Global Moderator

Offline

Posts: 2386
Silicon Valley

Re: Proper power settings for modulated RF signals
Reply #3 - Jan 8^th, 2007, 4:12pm

It is not calculating the transfer function per see. Rather you apply a signal and it computes a response for that signal assuming the circuit is linear. Normally people apply a 1V signal, in which case the output is scaled in such a way that it gives the transfer function directly.

It is not clear to me whether your receiver would care about the modulation on the blocker as I would not think that it would demodulate it. However, if it were me I would probably try to do some high-level simulations with a system simulator or behavioral models to determine if the modulation on the blocker is significant. In that case, using transient analysis should be fine. If you really want to do things at the circuit level, you could try using envelope simulation.

-Ken

Back to top

IP Logged

NKS New Member Offline Posts: 7 Seoul, Korea	Re: Proper power settings for modulated RF signals Reply #4 - Jan 8^th, 2007, 4:44pm Thank you very much, good ghost . NKS
Back to top	IP Logged