The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Simulators >> RF Simulators >> IIP3 analysis on a netlist with nport statement
https://designers-guide.org/forum/YaBB.pl?num=1076097546

Message started by amitava on Feb 6^th, 2004, 11:59am

Title: IIP3 analysis on a netlist with nport statement
Post by amitava on Feb 6^th, 2004, 11:59am

Hi,
I am facing a problem when doing Cadence Spectre simulation for IIP3 analysis on a netlist that has "nport" statements modeling the interconnects of a circuit. As an example I am giving the nport statement of a 2 port network that I have in my netlist below:
--------------------------------------------
simulator lang=spectre
model sp_data1 nport file="spNport1.data"
//Components
X1 (a1 a2 b1 b2 )sp_data1
-------------------------------------------

The file "spNport1.data" contains the s-parameters (2x2 matrix) of the interconnects (a1 to a2 and b1 to b2) as a 2-port network. The netlist has other elements like resistors, transistors, inductors and capacitors connected to each other to form an LNA topology. There are other "nport" statements too for other connectivities but I have given an instace with 2 nets connecting a few of the components.

When I simulate the entire netlist for doing the IIP3 analysis I get an error like this:
************************************
Error found by spectre during periodic steady state analysis `pss1'.
Distributed components and components with hidden state are not allowed
with this analysis -- analysis skipped.
nport

Analysis `pss1' terminated prematurely due to error.
**************************************
As pss and pac analysis are essential to do the IIP3 analysis, I am not sure how to avoid this error from happening and to determine IIP3 of a netlist that has "nport" statements describing the interconnects.

(I went thru the designers guide on how to avoid "hidden-state" errors in spectre but am clueless on how to avoid the error in this case)

Can anyone of you help me on this.

Thanks and Regards,
Amitava

Title: Re: IIP3 analysis on a netlist with nport statemen
Post by Andrew Beckett on Mar 5^th, 2004, 1:52am

You will need to set "interp=rational" on the nport
("Interpolation Method" on the edit parameters form in
the schematic), together with either ratorder (to indicate
the "order" it should use for the rational approximation when
producing a time domain representation. Alternatively
you can use relerr/abserr to control the accuracy
(see "spectre -h nport" for more info).

Currently only this approach for generating a time-domain
representation will work with spectreRF, because spline
and linear interpolation methods have hidden states
(as you've seen). In IC5141, all modes will work with
SpectreRF though (due out in June).

Andrew.

Title: Re: IIP3 analysis on a netlist with nport statemen
Post by CDR on Mar 24^th, 2004, 6:44pm

What should be the format of spNport1.data?Could you give examples?

Title: Re: IIP3 analysis on a netlist with nport statemen
Post by Andrew Beckett on Mar 28^th, 2004, 11:36am

The format can be spectre's own format (I think this
is covered in the documentation), touchstone or citifile
format. From IC5033 this can be selected on the instance
of nport itself, but there's also a translator "sptr" which
can be used to do the conversion (translator also
available in older versions).

Andrew.