PSS and HB vs. Transient having touchstone models from HFSS

Regarding:

HamidKhatibi said:

Hello every one

I have designed a 100GHz oscillator and in order to see how much power it can produce I run different analyses. The issue is that the output voltage is quite different when I use different analyses. The frequencies are similar but the amplitudes are totally different. I have used PSS with harmonic balance and also probe-based HB (these two simulation give very similar results). However when I use Transient, the amplitude of the node voltages is quite different (the frequency is very similar).

Which one can I trust?

Thank you very much,

Hamid

PS: I have couple of transmission lines whose model are coming from EM simulation in HFSS.

There could be a number of things going on.

If you have access to Cadence Online Support, please check out these articles:

• 20279103 MMSIM 13.1 Nport Overview and Guidelines

http://support.cadence.com/wps/mypoc/cos?uri=deeplinkmin:ViewSolution;solutionNumber=20279103

•  20158580 Tips for Using S-Parameters in Spectre and SpectreRF  (discusses troubleshooting)

http://support.cadence.com/wps/mypoc/cos?uri=deeplinkmin:ViewSolution;solutionNumber=20158580

• 20125630 Significant Improvements in S-parameter (nport) Simulation in MMSIM 12.1 and MMSIM 12.1.1  (particularly if using linear/spline interpolation)

http://support.cadence.com/wps/mypoc/cos?uri=deeplinkmin:ViewSolution;solutionNumber=20125630

 

1. What version of MMSIM are you using?  Typing the command in an xterm   spectre -W   will give this info.

2. What are your nport settings?  (Post a snippet from the netlist).

 

Best regards,

Tawna

Hi Twana.

Thank you  very much for your response. Unfortunately as a PhD student I have no access to those links.

The version is : 11.1.0.722.isr24

I have already read your useful note about the nport's settings and I have applied them in it.

Thank you very much for your helpful responses.

Hamid

The netlist is:

// Generated for: spectre
// Generated on: Dec  4 08:22:54 2014
// Design library name: 94GHz
// Design cell name: BJT_Osc94GHz_NoYs_DCCapVer5_TL4u30u
// Design view name: schematic
simulator lang=spectre
global 0
parameters Vdc2=880m Length=2.5 R3=1307 R2=51k R1=51k C1=4.88 C2=1.645 \
    C3=67.88 Q=50 Cs=44 RL=50
include "/research/afshari-new/hk558/Cadence/ST_BICMOS130/Corners/corners.scs"

// Library name: 94GHz
// Cell name: BJT_Osc94GHz_NoYs_DCCapVer5_TL4u30u
// View name: schematic
C42 (net2 0) capacitor c=C1*1e-15
C1 (net2 net023) capacitor c=1u
C40 (net020 0) capacitor c=Cs*1e-15
C44 (net016 0) capacitor c=C2*1e-15
C43 (net016 net2) capacitor c=C3*1e-15
Q5 (net20 net021 net020 0) npnvhsM2 l=Length*1e-6 nbc=4 nbb=6 nbe=3 mult=1 \
        mismatch=1 sh=1 lpe=0 dic_mdev=0 dib_mdev=0 soa=1
R22 (net2 net016) resistor r=R3
R23 (net2 0) resistor r=R1
R24 (net016 0) resistor r=R2
R21 (net020 0) resistor r=Q/(6.28*180*Cs*1e-6)
R19 (net08 net019) resistor r=RL
V9 (net17 0) vsource dc=Vdc2 type=dc
V8 (net08 0) vsource dc=1.6 type=dc
L21 (net17 net021) inductor l=10u
L25 (0 net020) inductor l=10u
I193 (net2 0) isource type=pulse val0=100.0u val1=0 rise=100f fall=100f \
        width=20p
NPORT8 ( net016 0 net021 0) nport interp=spline fmax=3000G fdelta=.1G \
        imptrunc=1e-6 thermalnoise=no usewindow=yes datafmt=touchstone \
        file="/research/afshari-new/hk558/sNp/TL4u30u_190u_HFSSDesign1.s2p" \
        romdatfile="TL4u30u_150u" dcextrap=unwrap hfextrap=linear \
        passivity=enforce pabstol=1e-06 interp=spline thermalnoise=no
NPORT9 ( net08 0 net019 0) nport interp=spline fmax=3000G fdelta=.1G \
        maxn=33 imptrunc=1e-6 thermalnoise=no usewindow=yes \
        datafmt=touchstone \
        file="/research/afshari-new/hk558/sNp/TL4u30u_550u_HFSSDesign1.s2p" \
        romdatfile="TL4u30u_170u" dcextrap=unwrap hfextrap=linear \
        passivity=enforce pabstol=1e-06 interp=spline thermalnoise=no
NPORT10 ( net20 0 net019 0) nport interp=spline fmax=3000G fdelta=.1G \
        maxn=33 imptrunc=1e-6 thermalnoise=no usewindow=yes \
        datafmt=touchstone \
        file="/research/afshari-new/hk558/sNp/TL4u30u_5u_HFSSDesign1.s2p" \
        romdatfile="TL4u30u_5u" dcextrap=unwrap hfextrap=linear \
        passivity=enforce pabstol=1e-06 interp=spline thermalnoise=no
NPORT5 ( net019 0 net023 0) nport interp=spline fmax=3000G fdelta=.1G \
        maxn=33 imptrunc=1e-6 thermalnoise=no usewindow=yes \
        datafmt=touchstone \
        file="/research/afshari-new/hk558/sNp/TL4u30u_130u_HFSSDesign1.s2p" \
        romdatfile="TL4u30u_360u" dcextrap=unwrap hfextrap=linear \
        passivity=enforce pabstol=1e-06 interp=spline thermalnoise=no
simulatorOptions options reltol=1e-3 vabstol=1e-6 iabstol=1e-12 temp=27 \
    tnom=27 scalem=1.0 scale=1.0 gmin=1e-12 rforce=1 maxnotes=5 maxwarns=5 \
    digits=5 cols=80 pivrel=1e-3 sensfile="../psf/sens.output" \
    checklimitdest=psf rawfmt=psfxl
tran tran stop=1n errpreset=conservative outputstart=.1n \
    write="spectre.ic" writefinal="spectre.fc" method=gear2only \
    annotate=status maxiters=5
finalTimeOP info what=oppoint where=rawfile
dcOp dc write="spectre.dc" maxiters=150 maxsteps=10000 annotate=status
dcOpInfo info what=oppoint where=rawfile
modelParameter info what=models where=rawfile
element info what=inst where=rawfile
outputParameter info what=output where=rawfile
designParamVals info what=parameters where=rawfile
primitives info what=primitives where=rawfile
subckts info what=subckts  where=rawfile
save R19:1 L21:1 V8:p
saveOptions options save=allpub

Hi Hamid, The first thing you need to do is upgrade your MMSIM software to MMSIM12.1.1 or later. There have been a number of notable improvements since then in the algorithms and interface. Best regards, Tawna

Hi Hamid, Please see this blog... It talks about the improvements to nport starting in MMSIM 12.1. http://community.cadence.com/cadence_blogs_8/b/rf/archive/2012/11/19/spectre-rf-mmsim-12-1-preview-of-coming-attractions

Hi Tawna

Thank you again.

Unfortunately it is not possible for me to upgrade MMSIM. What do you think about those different outputs (PSS hb and Transient)? (The output power are 2.5dBm and 5.7dBm which are seriously different and unfortunately the optimum load is totally different for them, so I have to know which one I can trust more.)

Thank you very much,

Hamid

Hi Hamid, I'm not sure if you are in the US or another country. Your University should have a program with Cadence where you can file a Case with Cadence Online Support. That way a Support AE (or myself) can take a look at your testcase. Without the actual data, it is difficult to say for sure - and I really hesitate to make a suggestion without seeing the s-parameter data. Many of the improvements in MMSIM12.1 and later releases are geared toward improving simulating s-parameters in transient analysis. For MMSIM 11.1, please see this article on how to set your nport parameters: http://community.cadence.com/cadence_blogs_8/b/rf/archive/2011/03/23/my-favorite-nport-settings But honestly, you really need to use a later version of code or have a Cadence Support AE take a look at this.

Hi Tawna

Actually I am a PhD student at Cornell university (Ithaca, NY) but we do not have this luxury here at Cornell!

Is there anything that I can send you so you might be able to help me?

BTW, I have read the mentioned article more than once and thank to your suggestions in that note I think that I have used the best settings that I could.

Thank you again,

Hamid

Hi Hamid, Good news.... Cornell is a Cadence University Program Member. Details are on the Cornell website: http://www.ece.cornell.edu/ece/about/cadence.cfm Please contact the person at the bottom of the website page or contact your department chair to gain access to Cadence Online Support. Unfortunately, I cannot help over the community forum. (Several Cadence AEs answer questions here "outside of our normal work responsibilities." To get specific help, you need to contact Cadence Customer Support.)

Thank you very much for all your support.

I had already asked people around and they did not know  this, even the professors. I will try that.

Thank you very much,

Hamid

If you are in the EE department (I also have names if you're in the physics dept). Contacts for Cornell University - EE are:
Licensing Professor: Carlos Otero
Primary Tech Liaison: Dora Abdullah
Secondary Tech Liaison: Doug McCool