Anomalies in the results of simulating frequency multipliers in ADS

[sunight]

Hello everyone.
I am designing a frequency multiplier using HFSS and ADS. However, the simulation curve in ADS produces a sudden change at a certain frequency point. After checking the snp file as well as the ADS settings, I really can't find the reason for this, can anyone explain the possible reasons? Below is my schematic as well as the simulation results (inputs/outputs are suspended microstrip lines converted to waveguides).

 

[volker@muehlhaus]

After checking the snp file as well as the ADS settings, I really can't find the reason for this, can anyone explain the possible reasons?

Does your SNP file cover the full frequency range from DC to the highest harmonics at 15*350 GHz = 5.25 THz?

I guess your data covers much less frequency range, and some extrapolation error happens when HB simulation needs data for the SnP blocks at the harmonics.

 

[sunight]

Does your SNP file cover the full frequency range from DC to the highest harmonics at 15*350 GHz = 5.25 THz?

I guess your data covers much less frequency range, and some extrapolation error happens when HB simulation needs data for the SnP blocks at the harmonics.

hello, I tried lowering the order to 3 (snp file ranges from 70 to 350GHz), I guess this shouldn't be out of range(I hope you can point out if my understanding is wrong), but the abrupt change in the simulation curve still exists.In addition, the position of this abrupt change shifts on the horizontal axis as the input power is changed.​

 

[volker@muehlhaus]

snp file ranges from 70 to 350GHz

This is a problem!
You really need to include the DC response in EM simulation, this is a must have!

Also, you should have data above 350 GHz, because HB simulation up to 350 GHz with order 3 requires data from DC to 3*350 GHz.

And finally, because your data for harmonics is incomplete or missing, check your interpolation settings for the SnP blocks in ADS. You can try the settings shown below.

 

[sunight]

This is a problem!
You really need to include the DC response in EM simulation, this is a must have!

Also, you should have data above 350 GHz, because HB simulation up to 350 GHz with order 3 requires data from DC to 3*350 GHz.

And finally, because your data for harmonics is incomplete or missing, check your interpolation settings for the SnP blocks in ADS. You can try the settings shown below.

View attachment 188209

Hello, after checking over the past few days, I realized that the cause of the problem is in the port settings in HFSS.
For the port on the diode, not selecting the normalization option results in the above problem, but I'm wondering why this is necessary.

 

[volker@muehlhaus]

This is an additional problem then, on top of the missing data.

SnP file require a constant reference impedance, and it seems that HFSS created something else.