pPar not works for intiger CDF parameters

 Andrew,

            I have created a inverter which passes paramers to bottom level. Then in top level schematic I have placed two inverters which has different widths (values entered in top level symbols). When I run LVS for a layout which has both inverters. Then I got some error's in LVS error report( looks like parameter errors, nets and devices). I mean is we cann't use two inverters in TOP level with different widths?

Thank you,

sarvani

Sarvani,

If things are set up correctly, this should work. It's been possible to do this for many years (I was using such parameterization 20 years ago as a designer). I suggest you contact customer support so that we can look at your data to find out what's gone wrong- some PDKs don't allow parameterization because they use CDF callbacks to generate derived parameters, and those derived parameters might be used for LVS - but I can't tell that without seeing the data (personally I think this is a badly designed PDK if this is the case, but it happens).

Regards,

Andrew.

Hi Sarvani,

I think that you have to do some other basic tests:

- first use same w everywhere without using pPar() , be sure you got LVS clean

- when LVS is clean try to play with pPar()

If the problem is still there, then it will be much easy for the others to find the problem if you can provide the "some error's in LVS"  and at least the netlist generated from schematic.

 Best Regards,

Marcel

Hi Andrew and Marcel. I had tested for improper cell (PMOS_5V). If I change the cell to NMOS then it works good. Looks like problem is solved. Thank you for your support.

Thank you,

Sarvani

Hi ,

 I will be grateful if someone can help me with a pcell simple issue. I have a lyout cell that instanciate a 1 parameter pcell. I want to move that parameter up in the hierarchy so that when I instanciate the cell and set the parameter, the pcell inside is updated accordingly. The same as the pPar("variable") in schematic.

Many thanks,

Mouadh

Hi Mouadh,

I'm assuming that the topmost PCell is a SKILL PCell?  When you instantiate the lower level PCell in the code you supply the parameter to that cell - the name of the parameter can be the same as the lower level cell, or entirely different, it's up to you.  So in your call to dbCreateParamInst or dbCreateParamInstByMasterName, you use that parameter in the list of parameters supplied to the instance you are creating, something like this:

dbCreateParamInstByMasterName(pcCellView "basic" "nand" "symbol" "inst1" 
list(0 0) "R0" 1 list(list("area" "float" myTopLevelAreaParam)) t )

In this case the lower level parameter is named "area" and the top level parameter is "myTopLevelAreaParam".  Hopefully this answers the question?

regards,

Lawrence.

Hi Mouadh,

Sorry, my example was mostly copied from the docs and shows a "float" parameter rather than integer, but hopefully you get the idea? You can even make the top-level parameter a string (often more flexible to have parameters as strings) and then used something like cdfParseFloatString() in the lower-level instance, or pcRound(cdfParseFloatString(myTopLevelAreaParam)), for example.

Hope this helps.

Lawrence.