S parameter model from manufacturer imported to ADS modified circuit performance

[Elijah Jacobson]

Hi,

I am a newbie in the RF field. I apologize for any stupid questions in advance.
I am working on a project in ADS and since the simulations gave me satisfactory results using ideal lumped components, I decided to replace these ideal components with commercially available models. So, I went to vendors' and manufacturers' sites and downloaded S parameter models for the components of my choice. Below is the snippet of circuit

I am trying to bring in the S parameter model for C41 1nf capacitor. I found a suitable component here at farnell. I went to manufacturer's website and downloaded the S parameter model here.
I have two questions.
1) The zip file downloaded has only series connection S parameter and as shown above, my capacitor is in parallel configuration in circuit. I cannot find shunt S parameters in website. Is it okay if I use the available series S parameters ?
2) Is there a difference between series and shunt S parameters ? if yes, can someone explain or lead me to a source where I can know more ? If there are no differences, then why does some manufacturer like Murata provide us with both shunt and series S parameters as seen here?

Now, since I did not have any other options, I took a leap of faith and imported the series S parameters for the capacitor and it changed the behavior of circuit in a bad way. So I am assuming that series and shunt S parameters are different. But I am not sure if this is happening because I am overlooking something or missing something. It will be nice to have an expert opinion on this.
Any help will be appreciated. Thank you very much for you time.

 

[volker@muehlhaus]

The series S-parameter data is correct for your use case. The difference to ideal element might be caused by physical length and internal signal path length, which means extra series L. What is the self resonance frequency of your 1nF capacitor, according to the S2P data? An ideal 1nF will be low impedance all the way to >100GHz, that is different from real components.

S-parameters measured in shunt configuration are something special, they can possibly provide more accuracy for very low impedance value. But that's not your issue here.

--- Updated Jun 22, 2023 ---

Edit:
I downloaded the S2P file and plotted effective series capacitance. You can see below that this capacitor has self resonance around 250 MHz, and C=1nF is only valid at frequencies much lower than this.

For real world components, SRF must be considered. Smaller C values have higher SRF.

You didn't mention your operating frequency, but most likely the 1nF value is not suitable for your needs. If you create an ADS testbench to plot impedance, you will see that this component (S2P data) isn't low impedance as you expected from the ideal cap.

 

[Elijah Jacobson]

Hi,
Thanks for replying. I think I understand it now. So what you are saying is that when picking components for RF applications, I need to look for components that have SRF above my operating frequency so that I get the desired value of impedance. right ?

You didn't mention your operating frequency,

Sorry about that. I am operating at 915 MHz.

Thanks again.

 

[volker@muehlhaus]

You need to realize that the effectice C value depends on frequency for the real component, and becomes inductive above SRF. So if you need a specific C value, you need to stay well below SRF, and you should use the component S2P data indeed.

If you just need the C to pass RF and the exact value isn't critical, you have more freedom to move towards SRF. The increase in C value isn't a problem in that case, as long as the device is low impedance.

Hope this makes sense?

 

[Elijah Jacobson]

That makes sense. Thank you so much !!