[SOLVED] Lumped Inductors in ADS

Hi,

I'm designing a simple RF lumped elements BPF in ADS. I simulated its S21 using ideal LC components and also used real inductors and capacitors models from the manufacturers web sites. Then i wanted to run a monte carlo analysis to find out the components tolerance impact on S21. I encountered the following problems:

1. When using inductors from ADS Lumped-Components menu, there is no inductor modeled with a resonance parameter (SRF), which is important when operating in frequencies near resonance.

2. When using "real" inductors' model from the manufacturers web sites (i assume this model contains the SRF parasitics), the inductance tolerance rating, in the "edit component parameters", appears between commas and can not be indicated as a parameter for statistical analysis (monte carlo) or yield analysis.

Either way may inductor models are not complete for proper simulation.

I'll appreciate your help in this matter.

Maybe you can use Microwave Office do the yield simulation.

Indeed, Microwave Office has these features but i wouldn't want using two CADs for doing a simple simulation unless i really have to, since most of my designs are already in ADS.
That is why i need to be 100% sure ADS is missing these simple features before deciding to split the simulation work.

RF lumped components are not recommended at all for hi-frequency problems. you'd rather use distributed elements.

ok, momo is right but ADS is accurate enough if you are around 1-2 GHz. Since you are using lumped elements you can try adding an ideal lumped element in series with your inductor and vary the value of that. I cant say if this will give you the best results but I think its worth a try.

Why don't you use your own models for lumped components ??
Try-for instance-Murata's impedance tool for lumped inductors and capacitors, find your equivalent circuit for each and then create your subcircuits with these elements as you wish.Put MC deviation limits,tolerances etc..
It will be more practical..

I always have the option to model the inductors myself but i don't think it is the most efficient way to work especially when you have many of them in your circuit.

If there is a similar way to work with inductors as in Microwave Office, i'll be glad to know about it. It gives you maximum flexibility, since you can vary any of the model parameters and the models contain SRF.

I'm surprised ADS doesn't offer similar model or simulation method, including SRF and inductance variance. For instance, if you take Murata models, although the model states the inductance tolerance, you can't integrate this parameter into Monte-Carlo.

I'll appreciate any additional information.

srull said:

I'm surprised ADS doesn't offer similar model or simulation method, including SRF and inductance variance. For instance, if you take Murata models, although the model states the inductance tolerance, you can't integrate this parameter into Monte-Carlo.

I'll appreciate any additional information.

Click to expand...

That's why I offer you create your own subcircuits that will consist of lumped ideal elements such as RLCK etc.So, all you should do find the equivalent circuit of a component and put your MC parameters and other stuffs..
Murata had a tool that is used to create an equivalent circuit of a capacitor or inductor but unfortunately they have pull it back from their website.It was used to see s-parameters of a C or L component with equivalent circuit.
You may still find this tool in somewhere..

BigBoss said:

That's why I offer you create your own subcircuits that will consist of lumped ideal elements such as RLCK etc.So, all you should do find the equivalent circuit of a component and put your MC parameters and other stuffs..
Murata had a tool that is used to create an equivalent circuit of a capacitor or inductor but unfortunately they have pull it back from their website.It was used to see s-parameters of a C or L component with equivalent circuit.
You may still find this tool in somewhere..

Click to expand...

The Murata tool offers very simple equivalent circuit (simple series RLC network) which is less accurate than using the combined serial/parallel netwok that i prefer to use.

I prefer using Microwave Office models instead of manually create less accurate models in ADS but need to understand first if ADS is really limited in this aspect.

srull said:

Microwave Office models instead of manually create less accurate models in ADS

Click to expand...

Microwave Office offers a wide range of inductor models, but these are really simple behind the scenes. In the element help, you can see the underlying schematic behind these Microwave Office inductor models. It only takes a few minutes to implement this in any other circuit simulator as a subcircuit. It's nice to have these models pre-defined, but is is very easy to implement them yourself if something is missing.

volker_muehlhaus said:

Microwave Office offers a wide range of inductor models, but these are really simple behind the scenes. In the element help, you can see the underlying schematic behind these Microwave Office inductor models. It only takes a few minutes to implement this in any other circuit simulator as a subcircuit. It's nice to have these models pre-defined, but is is very easy to implement them yourself if something is missing.

Click to expand...

When you use such a comprehensive simulation tool as ADS you can at least expect decent components' models. It is probably a limitation of ADS and it is either using Microwave Office, as suggested by tony_lth, or generate my own models as suggested by you volker_muehlhaus, by BigBoss, and by mkonca.

Thank you all for your suggestions and advices, i appreciate them very much.
Many thanks.