OCEAN script Problem

Hi Zheng,

This message appears for different reasons, sometimes due to bugs, or perhaps due to SKILL syntax problems in the code.  You can use the SKILL Lint checker to help identify syntax issues. If this does not help, I suggest that you may want to contact Cadence Customer Support for additional help.  Be sure to provide the subversion that you are using since some of these issues have already been fixed.

Hope this helps!

Regards,

Lawrence.

Hi,

Really cool thanks. The code used to be loaded well in another server and it is a simulation script for an LNA circuit. Right now I just paste the code and hopefully you are interested.

I provide the problematic sample.ocn code in two parts: Run-through part and error-warning part.

 The code can be successfully loaded the first half:

simulator( 'spectre )
design(  "./netlist")
resultsDir( "./schematic" )
modelFile(  "./mitllfdsoi0p18um/MP6_2p2/models/spectre/nom/mitllfdsoi0p18um.scs")
desVar(   "cw3" 10f )
desVar(   "cw1" 10f )
desVar(   "rw3" 1 )
desVar(   "rw1" 1 )
desVar(   "rls" 2 )
desVar(   "rlg" 7 )
desVar(   "rld" 6 )
desVar(   "rg" 0.5 )
desVar(   "Ls" .65n )
desVar(   "Lg" 2n )
desVar(   "Ld" 2.3n )
desVar(   "cds" 28.8f )
desVar(   "Cd" 160f )
desVar(   "Co" 120f )
desVar(   "cw4" "cw3" )
desVar(   "cw2" "cw1" )
desVar(   "rw4" "rw3" )
desVar(   "rw2" "rw1" )
Cnominal = desVar( "Cnominal" 0 )
Rnominal = desVar( "Rnominal" 0 )
R14 = desVar( "R14"  0)
      desVar( "C14" Cnominal )
R18 = desVar( "R18"  0)
      desVar( "C18" Cnominal )
R15 = desVar( "R15"  0)
      desVar( "C15" Cnominal )
R19 = desVar( "R19"  0)
      desVar( "C19" Cnominal )
R24 = desVar( "R24"  0)
      desVar( "C24" Cnominal )
R23 = desVar( "R23"  0)
      desVar( "C23" Cnominal )
R25 = desVar( "R25"  0)
      desVar( "C25" Cnominal )
R26 = desVar( "R26"  0)
      desVar( "C26" Cnominal )
desVar(   "frf" 5.6G )
desVar(   "df" 10M )
desVar(   "frf2" "frf+df" )
desVar(   "prf" -30 )
temp( 27 )
ResDir = "./sample"
resultsDir(ResDir)
sprintf( resultsFile "%s.txt" ResDir )
of = outfile(resultsFile)
startPoint = 2
stopPoint = 2
totalPoints = 1
resistorList = list("rls")

But when the second half below is loaded, the error came out". The second half is:

foreach( sweepVariable resistorList

         paramAnalysis(sweepVariable ?start startPoint ?stop stopPoint ?lin totalPoints)

     ;S-Parameter analysis
         analysis('sp ?ports list("/PORT1" "/PORT2")  ?start "400M"  ?stop "8G" 
        ?lin "601"  ?donoise "yes"  ?oprobe "/PORT2"  ?iprobe "/PORT1"  )

;1dB compression point and IIP3
;
;Periodic Steady State
         analysis('pss ?fund "frf"  ?harms "20"  ?errpreset "moderate" 
         ?param "prf"  ?start "-30"  ?stop "10"  ?lin "20"  )
         analysis('pac ?sweeptype "absolute"  ?start "frf2"  ?stop ""  ?maxsideband "2"  )

         paramRun()

         openResults(ResDir)
;selectResults('sp)

;;;;;5.6GHz
         freq = 5.6e+09

;Input Port Reflection (S11)
;measured @ 5.6G (lower s11 is better)
         s11 = value(dB20(aaSP(1 1)) freq)
;         s12 = value(dB20(aaSP(1 2)) freq)
;         s21 = value(dB20(aaSP(2 1)) freq)
         s22 = value(dB20(aaSP(2 2)) freq)
;minimum input reflection in frequency range simulated
;minreflection = ymin(dB20(aaSP(1 1)))
;other S-parameters

;Noise Figure
         noiseFigure = value(dB10(NF()) freq)
;noiseFigureMin = value(dB10(NFmin()) freq)

;Gain
;         transducerPowerGain = value(dB10(GT()) freq)
;         operatingPowerGain = value(dB10(GP()) freq)
         availablePowerGain = value(dB10(GA()) freq)
        
;selectResults('sweeppss_fd-sweep)
;         onedBcompPt = compressionVRI((v("/net016" ?result "pss_fd") - 0.0) '1 ?rport resultParam("PORT2:r" ?result "pss_fd") ?gcomp 1)
;selectResults('sweeppss_pac-sweeep)
         iip3 = ipnVRI((v("/net016" ?result "pac") - 0.0) '-2 '0 ?rport resultParam("PORT2:r" ?result "pac"))

;;;;;header
;         fprintf(of
;          "%s\ts11\ts12\ts21\ts22\
;\tnoiseFigure\ttransducerPowerGain\toperatingPowerGain\tavailablePowerGain\tonedBcompPt\tiip3\n" sweepVariable)
         fprintf(of "%s\ts11\ts22\tnoiseFigure\tavailablePowerGain\tiip3\n" sweepVariable)
;         ocnPrint( s11 s12 s21 s22 noiseFigure transducerPowerGain operatingPowerGain availablePowerGain onedBcompPt iip3 ?output of)
         ocnPrint( s11 s22 noiseFigure availablePowerGain iip3 ?output of)
        
         ) ; foreach(sweepVariable resistorList

drain(of)
close(of)

Thanks a lot before hand and after.

Steven.L

Hi,

I further tried another ocn file written with cadence analog env and it can be loaded well:

 ocnWaveformTool( 'wavescan )
simulator( 'spectre )
design(  "/nfs/users.labnet.st5/mnt/zl3140/simulation/lnaSimple/spectre/schematic/netlist/netlist")
resultsDir( "/nfs/users.labnet.st5/mnt/zl3140/simulation/lnaSimple/spectre/schematic" )
modelFile(
    '("/nfs/users.labnet.st5/mnt/zl3140/NEOCELLRUN/NeoCell/models/spectre/rfModels.scs" "")
)
analysis('ac ?start "1 "  ?stop "20 "  )
desVar(   "prf" -10     )
save( 'v "/net28" )
temp( 27 )
run()

By the way, the sample ocn codes in artist/spectre/*.ocn also bring error messages of "string/symbol" or similar stuff which are not supposed to appear.

 Right now I am comparing the good ocn above and previous error ocn to find some hints.

Steven.L

Hi Steven,

 One suggestion I can make is to insert debug messages at various locations, something like the following:

fprintf(stderr "DEBUG1: before foreach\n") ;;debug


foreach( sweepVariable resistorList
    fprintf(stderr "DEBUG2: before parametric analysis\n") ;;debug

     paramAnalysis(sweepVariable ?start startPoint ?stop stopPoint ?lin totalPoints)


    fprintf(stderr "DEBUG3: before S parameter analysis setup\n") ;;debug


      ;S-Parameter analysis
         analysis('sp ?ports list("/PORT1" "/PORT2")  ?start "400M"  ?stop "8G" 
        ?lin "601"  ?donoise "yes"  ?oprobe "/PORT2"  ?iprobe "/PORT1"  )
 ...

This way you can get useful information as to where the problem is arising; once the offending code has been isolated, it will be easier to track down and debug the cause.  By putting DEBUG in the print output and "debug" in the comment you have two chances to find these statements later when they need to be removed.

I hope that this helps you to find and solve the problem.

Regards,

Lawrence.

 Zheng,

By the way, this question may get better responses if posted to a more appropriate forum, i.e the Custom IC Design forum or even the Custom IC SKILL Forum

Regards,

Lawrence.