Editing a Netlist Before Simulation Starts

Karam,

Assuming you're talking about the "preRun" signal for ADE XL, then the answer is that this runs before netlisting, since usually the netlisting will be done in the background ICRP processes and preRun is triggered before any of this starts. Even if you have ignoreDesignChangesDuringRun set, this it would be before the netlisting gets triggered.

Probably anything that alters the netlisting would need to be done as part of the netlisting flowchart steps. Not sure this is a good idea - or why you really need to do this - but here's an article which explains how to have a trigger at the right point:

SKILL: How to execute a trigger at netlist time in both ADE L and ADE XL?

With ADE Explorer and Assembler there's a chance that the same netlist may be used for multiple points, so it depends on what exactly you're trying to "tweak" as to how successful this will be.

Andrew.

Hello Andrew,

Thank you for the quick reply. The link is helpful, I will try it out. I am asking because I would like to change the include statement in the netlist to add different sub-circuits to different simulation corners.

Would the link you provided apply to the GXL optimiser as well?

Your valuable time is highly appreciated.

Best regards,
Karam

KGh94 said:

Would the link you provided apply to the GXL optimiser as well?

Should do.

Why don't you change the model setup in each corner instead? That's what it's for - rather than hacking the netlist...

Andrew.

Hello Andrew,

Thanks for the suggestion. The main reason is because the netlister is struggling when some pcell parameter is changed. The callback is not applying the change in the netlist in a correct matter, causing simulations to crash. Hacking the netlist is a work-around to this problem.

Best regards,
Karam

Hi Karam,

Well, if it's really a pcell (e.g. a schematic pcell, where there's code which changes the structure and connectivity of the schematic), and it's a parameter that controls that which is being varied, then this is a known limitation of the "Parameterisation" flow (we understand the problem - it's just not that straightforward to fix). Not sure if you really mean a pcell or not or just a cell.

If it's not a pcell (a "parameterised cell" schematic) then this is something you should absolutely report to customer support.

Regards,

Andrew.

Hi Karam,

Well, if it's really a pcell (e.g. a schematic pcell, where there's code which changes the structure and connectivity of the schematic), and it's a parameter that controls that which is being varied, then this is a known limitation of the "Parameterisation" flow (we understand the problem - it's just not that straightforward to fix). Not sure if you really mean a pcell or not or just a cell.

If it's not a pcell (a "parameterised cell" schematic) then this is something you should absolutely report to customer support.

Regards,

Andrew.

Hello Andrew,

It is a schematic pcell, and it's the limitation you're talking about. The changes that need to be made in the netlist are known so hacking is the way these changes are going to be implemented. If a workaround that doesn't involve hacking the netlist exists, I would be open for suggestions. Otherwise there is no choice 

Best regards,
Karam