EM SIlmulation in ADS

[inass57]

Hi guys,

I have designed a basic RF-DC decoupling circuit for my amplifier on ADS using schematic, and my simulation were the one below:

And when i generate the layout to simulate it on momentum i get these results:



Here is the circuit if it might help:


Thank you in advance.

 

[BigBoss]

-Check the Substrate
-Your meshing seems a bit coarse, refine it.
-Choose appropriate EM Simulation Mode ( Microwave ).

 

[inass57]

-Check the Substrate.

I don't see a problem with my substrate definition:

-Your meshing seems a bit coarse, refine it.
-Choose appropriate EM Simulation Mode ( Microwave ).

I tried both, doesn't change anything

 

[volker@muehlhaus]

Circuit models are not exact. If you have done the EM simulation correctly, it will be more accurate than the microstrip circuit models.

Two details where you can improve the accuracy of your EM simulation:
- Include metal thickness in EM simulation, instead of using thin sheet where thickness is only used to calculate loss.
- Your port 3 might be a problem, if Momentum places the port across the entire edge. See my appnote: https://muehlhaus.com/support/ads-application-notes/edge-area-pins

 

[inass57]

I've used the edge PIN as was adviced in your answer, it did help but still not getting the same resuolt as in my schematic. and i still can't pin point the problem.

Here below the results of the simulation after i changed the pins:

 

[BigBoss]

But the last one is not too bad.It just needs a bit tuning.

 

[inass57]

Okay thank you, i didn't work a lot with Momentum, so i thought if my design is right i'm supposed to have the same results without editing my design.

 

[volker@muehlhaus]

but still not getting the same resuolt as in my schematic.

Schematic models are not perfect, that is why we use EM simulation for accurate results.

 

[inass57]

One last question: i tuned my design of the bias tee and wanted to use it to design the LNA, so i created a symbol for the layout of the bias Tee and integrated it on the schematic of the LNA as shown below, but the I probe shows a value of 3e-13 while before the bias (on the R2 resistance) i have 50µA. so my supposition is that the Bias Tess doesn't conduct current.
How can i remedie to this? on the Model options i did choose "S parameter data and currents or fields for post processing"


Thank you so much for your help.

- - - Updated - - -

Hi, i was able to solve it on my own, i didn't simulate my BIAS Tee at 0 GHz.
Have a nice day!