How could i get a Plot with a Xaxis in volt by the ocean script?

That's because you have selectResult('dcOp) and it should be selectResult('dc). What you're accessing is the dc operating point, which doesn't change throughout the sweep - that's the initial operating point at the beginning of the sweep, and what you actually want is the result of the dc analysis itself.

Regards,

Andrew.

Great, thank you Andrew.

By the way, i have a other question about variable in an expression of monteCarlo analyse.

 Vdd_half=0.5 for example

 "Normal" simulation:

  analysis('tran ?stop TempoSimu  )

  temp( myTemp )
  run()
  selectResult( 'tran )

  Delay_WL_S_rise = delay(clip(v("/WL" ?result "tran-tran") 2.8e-08 3.6e-08) Vdd_half 1 "either" clip(v("/S" ?result "tran-tran")     2.81e-08 3.21e-08) Vdd_half 1 "either" 0 0 nil nil)

i obtain the result, it's work ...

-------------------------------------------

 in the "monteExpr" when i do the same thing, its impossible, i need to put the right value directly. Have you got an idea (syntax) to allow me to put the variable instead of the value?

 MonteCarlo simulation:

monteExpr( "Delay_WL_S_rise" "Delay_WL_S_rise=delay(clip(v(\"/WL\" ?result \"tran-tran\") 2.871401e-08 3.685221e-08) 0.50 1 \"either\" clip(v(\"/S\" ?result \"tran-tran\") 2.871401e-08 3.685221e-08) 0.50 1 \"either\" 0 0 nil nil)" )

Thank you for your answer...

Hi Axel,

I think (I've not tested this), you should be able to do:

desVar("Vdd_half" 0.5)

and then in the monteExpr call VAR(\"Vdd_half\") wherever you have the 0.5 right now.

Regards,

Andrew.

Andrew,

There certainly was an issue with using VAR() in Monte Carlo simulations which was fixed in IC614 ISR1. So maybe your grief will be resolved by that?

Andrew (B)

 Well, so i declare Vdd_half as "desvar" and after, for using it in the monteExpr, i need to use VAR() or pv() ?

but after i'm lost for that... sorry :)  i have unbound variable \"Vdd_half\" with VAR(), pv() or without nothing.

on ocean ref or skill guide they don't speak about these function... how can i use them?

thank you both of you

Axel

Hi Axel,

Here's a very simple OCEAN script I have using VAR and pv() approaches:

ocnWaveformTool( 'wavescan )
simulator( 'spectre )
design(  "/export/home/users/andrewb/simulation/rcac/spectre/schematic/netlist/netlist")
resultsDir( "/export/home/users/andrewb/simulation/rcac/spectre/schematic" )
path( "./models" )
modelFile(
    '("res.scs" "")
)
analysis('tran ?stop "100n"  )
desVar(   "sup" 3       )
desVar(   "rval" 1k     )
desVar(   "cval" 1p     )
temp( 27 )
monteCarlo( ?numIters "100" ?startIter "1"
    ?analysisVariation 'process ?sweptParam "None"
    ?sweptParamVals "27" ?saveData nil
    ?nomRun "yes" ?append nil
    ?saveProcessParams t
    ?samplingMethod 'LHS)
monteExpr( "RT" "riseTime(clip(VT(\"/net5\") 8e-09 1e-07) 0 nil VAR(\"sup\") nil 10 90 nil \"time\" nil nil)" )
monteExpr( "RVAL" "VAR(\"rval\")" )
monteExpr( "RVAL2" "pv(\"top-level\" \"rval\" ?result 'designParamVals)" )
monteRun()

Both work. In this case I have the variable "sup" (which is constant) which is used within the RT measure. Then I have rval (which has a statistical distribution on it, inside the res.scs model file). You can see two different ways of accessing that variable, with VAR and pv(). Both give a distribution as in the attached picture.

The pv function is definitely covered in the OCEAN Reference manual (chapter 7, Data Access Commands in IC5141). Didn't check the VAR() function, but it corresponds to the var button in the info tab (in wavescan calculator), or the var button in the AWD calculator.

Regards,

Andrew.

In my web browser, the attachment picture in the last post is chopped off and only shows two out of the three subwindows - if you click on the attachment name though you'll see all three monteExpr histograms.

Regards,

Andrew.

Hi Andrew,

Thank you for your help. I was trying to do the same thing for my application,

ie: put Vdd_half as a desVar and put VAR(\"Vdd_half\") in the monteExpr.

I think i've put exactly the same thing that on your script for your variable 3 and.........ERROR !!

Whereas when i put 0.55 instead of VAR(..) it works ! i don't understand how it works on your test. Do you see something wrong?

Here is the main part ofmy ocean script:

---------------------------------------------------------------------------------------------

ocnWaveformTool( 'wavescan )
simulator( 'spectre )
design(     "/auto/ax/simulation/Timing_tb/spectre/schematic/netlist/netlist")
resultsDir( "/auto/ax/simulation/Timing_tb/spectre/schematic" )
definitionFile(
    "models.scs"
)
analysis('tran ?stop "80n"  )

desVar("Vdd_half" 0.55)
temp( 27 )
monteCarlo( ?numIters "2" ?startIter "1"
    ?analysisVariation 'processAndMismatch ?sweptParam "None"
    ?sweptParamVals "27" ?saveData t
    ?nomRun "yes" ?append nil
    ?saveProcessParams t
)

monteExpr( "Delay_WL_S_rise" "delay(clip(v(\"/WL\" ?result \"tran-tran\") 2.871401e-08 3.685221e-08) VAR(\"Vdd_half\") 1 \"either\" clip(v(\"/S\" ?result \"tran-tran\") 2.871401e-08 3.685221e-08) 0.55 1 \"either\" 0 0 nil nil)" )
monteRun()

 ---------------------------------------------------------------------------------------------------------------------

ERROR SPECTRE 8003 error evaluating ocean expression delay(clip(v(\"/WL\" ?result \"tran-tran\") 2.871401e-08 3.685221e-08) VAR("Vdd_half") 1 \"either\" clip(v(\"/S\" ?result \"tran-tran\") 2.871401e-08 3.685221e-08) 0.55 1 \"either\" 0 0 nil nil)

----------------------------------------------------------------------------------------------------------------------

thanks

Axel