[SOLVED] How to design variable resistors R0 to represent the frequency-dependent loss in ADS

Seems like it just wants a wee bit of series L?

Use equation based s-parameter blocks.
Define first s-parameters matrix of a series resistor then enter these values into equation based blocks with the variable "freq". Otherwise it will not recognize.

BigBoss said:

Use equation based s-parameter blocks.
Define first s-parameters matrix of a series resistor then enter these values into equation based blocks with the variable "freq". Otherwise it will not recognize.

Click to expand...

Thank you very much for your response, until now I updated the equivalent circuit as follows, but still I cannot reach the proper data. Do you have any idea where the problem is? Also, i do not know how to introduce the frequency as the variable "freq". I tried to but the software does not accept it.
any suggestion appreciated.

dick_freebird said:

Seems like it just wants a wee bit of series L?

Click to expand...

I do not know, would you please explain more?

@RealAEL Do you have any idea?

S-parameters matrix of a series resistance ;

Here Z=R in your case.
Then use equation based s-parameter block like this.

then you will see that s-parameters are not constant versus frequency. (If the resistance was constant, s-parameters should have been constant too but they aren't due to frequency dependence of the resistance)

BigBoss said:

S-parameters matrix of a series resistance ;
View attachment 180869
Here Z=R in your case.
Then use equation based s-parameter block like this.
View attachment 180870
then you will see that s-parameters are not constant versus frequency. (If the resistance was constant, s-parameters should have been constant too but they aren't due to frequency dependence of the resistance)
View attachment 180872

Click to expand...

Thank you very much for your guidance I did what ever you mentioned in the previous comment as follows based on my current device and here is my Equivalent circuit:



S21, S11 are my experimental results and S65, S55 are for the simulation



I've got a good verification between S21 and S65, However the S55 does not seems good and it starts from -10dB
So based on the article I changed the frequency loss resistance depedance as follows: I eleiminated R=50

then the slope of S55 changed but still it starts from -10dB. It seems the variable Rlossy has lots of effect on our measurments and by changing the formula i even needed to change the capacitance C4 for validation between simulation and experiment results.




Would you please guide mewhich one is current and where did I made a mistake because S55 should start from zero.

It was my own example, not real/your one. You have to define your own formulae by yourself. It might
be anything as you desire. I dont know the process that you've been working on.

I would model this as S1P data block, using Z parameters in real/imag format for easy data entry.

# GHz Z RIR 1.0
0.00 5 0
1.00 6 0
2.00 7 0
10.0 15 0

In ADS that S1P block is connected as a series element using both terminals (signal and ground).

edit: Typo in the file header, it should be

# GHz Z RI 1.0
0.00 5 0
1.00 6 0
2.00 7 0
10.0 15 0

thundrous said:

I do not know, would you please explain more?

Click to expand...

If it's a real resistor then it has real inductance and perhaps
modeling it that way gets you where you need to go. The
stated behavior seems consistent with that (though whether
a simple lumped, or more fine grained network suits your goals
I could not say)