Transient Simulation with NPORT from analogLib

My guess is that you're not using a Cadence version of analogLib, but one that has been adapted for another simulator (you may even be using a different simulator - you didn't say). The reason I suspect this is because the nport form doesn't look like it normally does nowadays - it usually looks like this (and has done for some time - your version is recent enough):

Can you go to analogLib in the library manager, and do Right Mouse->Properties over the name of the library and check the read path. It should be coming from within the ICADV123 installation. Maybe it has been generated via some script which updates the CDF to suit another simulator, and this has messed up the CDF so that it doesn't match the device? Normally the nport has a common reference node (at the bottom) and then each port terminal is single ended rather than differential - it looks as if you have a different (non-standard) netlisting procedure which doesn't know how to deal with this - or your nport doesn't have that pin, and it's confused.

Anyway, I'd start with checking whether you have a non-Cadence analogLib in your cds.lib.

Regards,

Andrew.

Andrew,

As always you figured it out at first shot. We had the following readpath in our cds.lib. I dnt know why.

/data/cad/Cadence/IC6/615_HF506/tools/dfII/etc/cdslib/artist/analogLib

Is this the old analogLib?

So thats why I didnt see a common terminal and also my input ports were all differential. And to answer your question, I am using ADE L for my simulations.

Now we have changed it to the path you mentioned, in our ICADV installation and that error is gone.

Thanks a lot for your help and support. I am not sure what we would do without this forum and especially you.

Thanks for making our lives easier.

Have a nice day.

Regards,

Amit

Hi Amit,

Yes, it's an old analogLib - not one for a different simulator, just one from an old IC version. The netlist procedure used comes with the IC tools themselves, so clearly there's an inconsistency between the symbol/CDF and the netlist procedure.

Anyway, glad it's solved and happy to help!

Regards,

Andrew.

Hi Andrew,

I am a little bit confused on how to to connect the nport in the case where I have a 2-port transmission line s-parameter file which is excited by a port on one side and loaded by a 1pF capacitor on the other side.

As shown in the figure there are different ways of connecting the nport instance,

Can you please explain which is the correct configuration and why?

You should really have started a new thread for this (see Guidelines for the RF Design Forum), but a quick answer.

B & C are identical - they will netlist the same way. In essence the common port of an nport means that all "minus" sides of each differential port are connected together, and so they are all connected to ground.

The top is likely to be a problem, since there's no ground defined for the simulator - everything is floating. So that means there's no reference. It rather depends however on the s-parameters - usually the ports are referred to something - usually some local ground signal - so I'd say B or C are more commonly used (it may not be gnd! you connect to, but it would commonly be some reference node which ultimately has a path to the simulator ground node).

Andrew

Thank you Andrew for the quick response. 

Dear VysakhK,

VysakhK said:

Can you please explain which is the correct configuration and why?

One added comment - to which I do not know the answer - is how the S-parameter file was measured. Specifically, in many cases, the ports defined for an S-parameter measurement are referenced to a standard resistance (typically 50 or 75 ohms). In your configuration B, n1 and n2 may be have been measured using 50 ohm or 75 ohm  port impedances. Hence, if you simulate the circuit using the measured S-parameters, the response using your network B may not be as expected since the port impedances of n1 and 2 are zero ohms.

You may want to check the measured S-parameter file to determine the its port impedances and choose your circuit topology to match it.

Shawn

m1 and m2 are not ports, they are the ground references of ports p1 and p2.