AnsweredAssumed Answered

HB2Tone results differ from HB2Tone_sweptfreq

Question asked by MURPHY294 on May 12, 2006
Latest reply on May 15, 2006 by MURPHY294
I have two simulations based on the amplifier design guide's HB2TonePAE_2swp template.

Sim1 is virtually identical to the template.  The HB simulation controller is set for FFT over sampling of 4, Restart=yes, and solver set to auto.  The sweeps are:

The resulting Vload when the RFfreq=3.2 GHz and fspacing is 2.0 kHz (so F1= RFfreq-fspacing/2 and F2=RFfreq+fspacing/2) is:

                                      freq                        Vload                        
Param1=0.005, Param2=1.560E-4                                                               
                                              0.0000 Hz           0.000 / 0.000
                                              2.000 kHz        4.110E-9 / 180.000      
                                              4.000 kHz       8.952E-17 / 169.412        
                                              3.200 GHz       3.227E-17 / -3.307      
                                              3.200 GHz        3.756E-13 / -51.710        
                                              3.200 GHz         8.759E-5 / 126.145      
                                              3.200 GHz         8.759E-5 / 126.145      
                                              3.200 GHz       6.493E-13 / -50.377        
                                              3.200 GHz       1.405E-17 / 75.722      
                                                 ...                  ...

Sim2 builds on sim1 but now the RFfreq is swept and the tones remain F1=RFfreq-fspacing/2, F2=RFfreq+fspacing/2.  The HB simulation controller is set exactly the same except for the fact that no sweep is set within the controller.  The configuration is thus:


                                  freq                       vs(Vload,freq)              
Param1=0.005, Param2=1.560E-4, RFfreq=3.200E9                                 
                                        0.0000 Hz            0.000 / 0.000
                                       2.000 kHz         4.162E-9 / 180.000      
                                        4.000 kHz        4.014E-17 / 179.208      
                                       3.200 GHz         1.436E-17 / -136.746      
                                       3.200 GHz         3.741E-13 / -52.303      
                                       3.200 GHz          8.835E-5 / 126.205      
                                       3.200 GHz          8.835E-5 / 126.205      
                                       3.200 GHz         3.741E-13 / -52.301      
                                       3.200 GHz         2.287E-17 / 23.131      
                                         ...                   ...
I have listed Vload vs freq here because the inner-most dependency is harmindex.  This listing just confirms that the indices correspond to the correct frequencies.  I was expecting the results of these two simulations to be the same but comparison of the listings reveal that the Vloads often differ in phase and thus magnitude.  While the differences are not extreme, my concern is that I am making a fundamental error someplace.

I have read a number of support documents and manuals that discuss swept HB and suggestions regarding nested sweeps and most of these concern convergence issues.  While I don't think this is a convergence problem, to my understanding I am following those guidelines.

I would like to know if anyone has experienced similar results or if you know what the problem may be.