Difference of S-parameters between Cadence Spectre and Keysight ADSsim
Cadence Spectre evaluates Y-parameters first internally.
Then it converts Y-parameters to S-parameters using reference impedance values.
Some simulator has option to choose Y or Z parameters.
On the other hand, Keysigh ADSsim does not use Y or Z parameters to calculate S-parameters.
Keysigh ADSsim calculate S-parameters directly by power wave formulation.
Power wave is defined by voltage, current an reference impedance.
Assume 2-port DUT where Y-parameters don't exist.
For example, shunt element.
Cadence Spectre can not give S-parameters.
Keysigh ADSsim can give S-parameters.
I don't think so, otherwise it would be great shame for Cadence.
Which version of spectre do you use but it computes s-parameters of an shunt inductor.
How about series element which does not have Z-parameters.
It's as expected..
This issue was filed as PCR(Product Change Request)-464911
"Systems which does not have a valid Y matrix give arithmetic exception"
Thanks for confirming.
I also confirmed.
Now this issue has been fixed in Cadence Spectre.
"sp_shunt.scs"
simulator lang=spectre
global 0
PORT1 (Vin 0) port r=50 num=1 type=dc mag=1
PORT2 (Vin 0) port r=50 num=2 type=dc
Rx (Vin 0) resistor r=50
//Rx (Vin 0) resistor r=50/2/(sqrt(2)-1)
heppoko options reltol=1e-5 vabstol=3e-8 iabstol=1e-13 temp=16.85 \
tnom=27 scalem=1.0 scale=1.0 gmin=1e-12 rforce=1 maxnotes=5 maxwarns=5 \
digits=5 cols=80 pivrel=1e-3 ckptclock=1800 \
sensfile="../psf/sens.output" rawfmt=psfascii
aho noise values=[100M] annotate=status oprobe=PORT2 iprobe=PORT1
boke sp values=[100M] annotate=status donoise=yes oprobe=PORT2 iprobe=PORT1
kasu ac values=[100M] annotate=status
hyottoko options save=allpub
HEADER "PSFversion" "1.00" "BINPSF creation time" "1564453319" "simulator" "spectre" "version" "17.1.0.307.isr6" "date" "11:21:59 AM, Tue Jul 30, 2019" "design" "(no title)" "analysis type" "sp" "analysis name" "boke" "analysis description" "S-Parameter Analysis `boke': freq = 100 MHz" "xVecSorted" "unknown" "tolerance.relative" 1.00000e-05 "mode" "ss" "number of ports" 2 "PORT2:r" 50.0000 "PORT2:num" 2 "PORT2:units" "Ohm" "PORT1:r" 50.0000 "PORT1:num" 1 "PORT1:units" "Ohm" TYPE "sweep" FLOAT DOUBLE PROP( "key" "sweep" ) "S-Param" COMPLEX DOUBLE PROP( "key" "SP" ) SWEEP "freq" "sweep" PROP( "sweep_direction" 0 "units" "Hz" ) TRACE "s22" "S-Param" "s12" "S-Param" "s21" "S-Param" "s11" "S-Param" "cy22" "S-Param" "cy12" "S-Param" "cy21" "S-Param" "cy11" "S-Param" VALUE "freq" 1.00000e+08 "s22" (-0.333333 0.00000) "s12" (0.666667 0.00000) "s21" (0.666667 0.00000) "s11" (-0.333333 0.00000) "cy22" (0.0200000 0.00000) "cy12" (0.00000 0.00000) "cy21" (0.00000 0.00000) "cy11" (0.00000 0.00000) END
That is why it is important to specify software version when discussing such issues. "Now" is not a useful specification.
See simulation result.
You can see version which I use.
However I don?t know in which version this issue was fixed.
I'm using Spectre 17.10.627 and it seems working..
But I personally don't like Spectre in RF simulation.You can even not attribute the frequency on Smith Chart to see impedance/admittance at particular frequency..Since 20 years..
My favorite one is Keysight GoldenGate. It's marvelous..
Right.
However this is not issue of Cadence Spectre.
This issue is for Cadence ViVA.
Simulation performance of current Cadence Spectre is fairly good even for multi-tone HB Analysis.
http://www.designers-guide.org/Forum...1170364605/2#2
https://designers-guide.org/forum/Ya...num=1205223090
https://designers-guide.org/forum/Ya...num=1162921214
It is very amazing.
I think no one use Keysight GoldenGate and even Keysight almost abandon promotion of GoldenGate anymore.
Keysight GoldenGate is almost EOL.
https://designers-guide.org/forum/Ya...m=1268404959#7
No, Goldengate is still promoted..
https://www.keysight.com/en/pd-29980...6&cc=US&lc=eng
This version will be compatible with ADS2020 and IC6.1.8.
There is no thread about Goldengate in this forum.
There is no reference of Goldengate in foundary document.
I foresee EOL of Goldengate in near future.
I often think this forum is most popular among the lazy students, people trying to find someome to solve their homework. There's few power users here .. not exactly the community to discuss advanced EDA stuff.