How to fix this: { *Error* eval: undefined function } on cadence IC5.1.4.1

Status
Not open for further replies.


So I just replace subckt to inline subckt in this file like the previous time.
But it didn't work like i told you.
I still don't understand why we insert "inline"?
Thank you.
 

KimPhan said:
So I just replace subckt to inline subckt in this file like the previous time.
But it didn't work like i told you.
I still don't understand why we insert "inline"?
Click to expand...
Wrong.
Surely see and understand what you have to change.
 
Last edited:

I am not convinced that the netlist is the problem. You'd
like to believe that the PDK is self-consistent and doesn't
netlist library devices in a busted way. Until you can say
-which- function is undefined and how it was supposed to
be declared, I don't think you can expect any particular
blind stab to work.

I'd try to find the function (have you yet parsed the CDS.log
file?) and then grep around for it and see if a whole file or
hierarchy has failed to load, possibly for reasons of munged
install or bad project initialization scripting.
 



I have already tried to replace subckt into inline subckt like your advice but it still didn't work.
I took another subckt in the PDK "nfet" and did simulation. The result is OK like the picture below. And in the model description of nfet,
it just have "subckt", not " inline subckt".


View attachment nfet_model.txt

So I don't think this solution is right. Do you have any suggestion?
Thank you.
 

Wrong.
Necessary task is not only adding "inline" word.
Surely see.
Necessary changes are only Three lines for each model.
 

pancho_hideboo said:
Wrong.
Necessary task is not only adding "inline" word.
Surely see.
Necessary changes are only Three lines for each model.
Click to expand...


Sorry for my carelessness!
But I edited the file of model description like this for each model:

inline subckt ln121_hicum (3 2 1 9)
ln121_hicum (3 2 1 9) mdn_hicum trise=temprise ( replace for "q 3 2 1 9 mdn_hicum trise=temprise")
ends ln121_hicum

But the result is unchanged. I think we cannot edit this file because it used to be run well before. But after a long time, I reinstall this
PDK and it have this error. So I think it is not the right method to fix this error.
 

KimPhan said:
So I think it is not the right method to fix this error.
Click to expand...
No.
Unless speciall callback function is invoked, "inline subckt" is necessary for annotating DC operation informations.

Do you surely activate modified model file ?
I suspect you still use original model files.
 

pancho_hideboo said:
No.
Unless speciall callback function is invoked, "inline subckt" is necessary for annotating DC operation informations.

Do you surely activate modified model file ?
I suspect you still use original model files.
Click to expand...


I edited the file. Then I saved it. And I restarted the server. then I tested the PDK and it was not like we expect.
 

See attached figure.
I can annotate DC Operation Point Information, if I modify model file as "inline subckt" correctly.
On the other hand, I can not annotate any DC Operation Point Information if I use your original model files which use "conventional subckt".

Here I use "analogLib/npn" as Symbol for BJT.

Code: 
// Generated for: spectre
// Generated on: Oct  5 13:31:26 2017
// Design library name: My_RFDE_Test
// Design cell name: test_KimPhan_npn_model
// Design view name: schematic
simulator lang=spectre
global 0

include "KimPhan_npn_model.scs" section=NO_RTH
include "KimPhan_npn_model.scs" section=NOM
include "KimPhan_npn_model.scs" section=CIRCUIT

// Library name: My_RFDE_Test
// Cell name: test_KimPhan_npn_model
// View name: schematic
V1 (net5 0) vsource dc=1 type=dc
V0 (net1 0) vsource dc=600.0m type=dc
Q0 (net5 net1 0 0) ln121_hicum
simulatorOptions options reltol=1e-3 vabstol=1e-6 iabstol=1e-12 temp=25.0 \
    tnom=25.0 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 
dcOp dc write="spectre.dc" save=allpub maxiters=150 maxsteps=10000 \
    annotate=status
dcOpInfo info what=oppoint where=rawfile
designParamVals info what=parameters where=rawfile
saveOptions options save=selected saveahdlvars=all

pancho_hideboo said:
No.
Unless speciall callback function is invoked, "inline subckt" is necessary for annotating DC operation informations.
Click to expand...
See attached document.

You might miss to load some skill function in your "~/.cdsinit" or "./.cdsinit".

Anyway, enter load "cdfdump.txt" in CIW.
Then show me generated file, "sbc18_npn.cdf".
 

Attachments

  • test_KimPhan_npn_model.png
    21.5 KB · Views: 107
  • cdfdump.txt
    101 bytes · Views: 57
  • 171005-205001.png
    50.8 KB · Views: 79
Last edited:
Status
Not open for further replies.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…